Key Signals

• PE-backed AI consulting ventures (DeployCo at $10B, Anthropic/Blackstone at $1.5B) are capital-structure confessions: the model alone cannot generate the enterprise revenue these companies need. The guaranteed 17.5% return to PE sponsors is a fixed-yield credit instrument, not a technology bet.

• Model commoditization is accelerating faster than Moore’s Law (50x annual inference cost decline), but a 643x pricing spread between commodity and frontier tiers allows consulting intermediaries to steer carriers toward expensive contracts when 85% of workloads could run at 1/50th the cost.

• Every major carrier AI deal targets the same thing: optimization of existing workflows through existing data. None address sensor integration, edge inference, or the Physical AI data layer where new insurance products will be designed.

• The edge AI chip market is scaling toward $57.8 billion by 2034. Emerging-market insurers are already operating sensor-to-payment architectures (weather stations to M-Pesa, telematics to usage-based pricing) that have no natural entry point for the LLM consulting layer.

• The execution gap is real: Path 1 deals consume the budget, engineering bandwidth, and board attention that Path 3 (Physical AI) would require. The trap is not that optimization is wrong. It is that optimization crowds out transformation.

1. The Confessions

On May 4, 2026, OpenAI finalized a $10 billion deployment venture called DeployCo with TPG, Brookfield, Bain Capital, Advent International, and fifteen additional private equity partners. The same day, Anthropic announced a parallel venture with Blackstone, Hellman and Friedman, and Goldman Sachs, capitalized at $1.5 billion. Zero investor overlap between the two ventures. Both announced within 36 hours.

These are not product launches. They are capital-structure confessions: the model alone cannot generate the enterprise revenue these companies need. They are also distribution confessions. Neither OpenAI nor Anthropic controls the surfaces through which enterprises access their products. Approximately 60% of OpenAI’s enterprise revenue flows through Microsoft Azure. Seventy-five percent of Anthropic’s API revenue flows through indirect channels: Amazon Bedrock, Google Vertex, and tools like Cursor. Anthropic’s CFO has acknowledged that partnerships are central to the company’s distribution strategy. When three-quarters of your revenue depends on platforms controlled by companies building competing models, a PE-backed consulting venture is not a growth strategy. It is a manufactured distribution channel.

DeployCo is a Delaware LLC led by OpenAI COO Brad Lightcap. OpenAI committed $500 million in initial equity with a $1 billion option to scale. The PE sponsors contributed $4 billion. The structure includes a guaranteed minimum annual return of 17.5% to PE backers over a five-year commitment -- more than double the industry-standard 8% preferred return. OpenAI retains control through super-voting shares. The PE sponsors have no governance power over model training or strategy. The investors backing a $10 billion AI deployment venture have no say in how the AI works. They have a guaranteed return. Section 6 examines what that return structure means for the enterprises on the receiving end.

Read that again. The investors backing a $10 billion AI deployment venture have no say in how the AI works. They have a guaranteed return. This is a fixed-yield credit instrument wearing a technology venture’s clothes.

Anthropic’s venture follows a different governance structure but identical logic. Public financial disclosures from the venture’s launch confirm the equity breakdown: Anthropic, Blackstone, and Hellman and Friedman each committed $300 million as individual anchor investments. Goldman Sachs and General Atlantic each committed $150 million, with GIC and additional participants bringing total capitalization to approximately $1.5 billion. The venture targets heavily regulated, document-dense financial verticals: investment research, underwriting analysis, compliance monitoring, fraud investigations. Anthropic reports that 40% of its top 50 customers are financial institutions, a figure disclosed at its “The Briefing: Financial Services” event in New York on May 5, 2026.

Both ventures deploy Forward Deployed Engineers at $200,000 to $300,000 total compensation into enterprises that cannot operate the software without them. DeployCo immediately acquired Tomoro, an applied AI consulting firm, bringing approximately 150 FDEs and active accounts including Tesco, Virgin Atlantic, Mattel, and Red Bull. The FDE model is not new. Palantir pioneered it a decade ago. Palantir’s stock has returned roughly 640% since its 2020 public listing, driven significantly by this model. The economics work for the vendor because the enterprise cannot evaluate whether the deployment succeeded. It is a credence good: the buyer lacks the expertise to assess quality, so the seller defines the terms of engagement and renewal.

The market understood the signal within days. On May 12, Indian IT outsourcers dropped sharply as investors priced in the structural threat: TCS dropped 3.8%, HCLTech fell 4.1%, Infosys fell 3.1%, and the Nifty IT index lost 7% over the following four trading sessions. The index had already shed more than 40% from its December 2024 peak. The market read these ventures as a structural claim that human deployment engineers, not autonomous AI, will be the revenue engine for enterprise AI adoption.

Both companies launched PE-backed consulting ventures within 36 hours. That timing is not coincidental. It is a coordinated market signal that neither company’s model revenue alone can sustain the growth trajectory their valuations require.

2. What Carriers Are Buying

The insurance carrier AI deal landscape is large, growing, and remarkably uniform in what it purchases.

Nationwide committed $1.5 billion to technology modernization through 2028, with 20% allocated to scaling AI. The deployments target underwriting guideline analysis, commercial submission summarization, and service request processing. State Farm joined as a launch partner for OpenAI’s Frontier platform, building what it calls “AI coworkers” for agents and claims. Liberty Mutual deployed LibertyGPT to 45,000 employees through OpenAI, achieving approximately 25% adoption and saving an estimated 17,000 administrative hours per week. Travelers equipped roughly 10,000 engineers and analysts with personalized Claude assistants; its claims contact center staff has been reduced by a third. AIG built an enterprise ontology integrating more than four million data points through Palantir Foundry and launched a Lloyd’s syndicate with Amwins and Blackstone with $300 million in initial stamp capacity.

In Europe, the numbers are larger. Allianz posted record operating profit of EUR 4.5 billion in Q1 2026 while registering more than 900 AI use cases; employees built 30,000 custom AI agents on the AllianzGPT platform serving 60,000 workers. Generali backed its “Lifetime Partner 27” strategy with an approximately EUR 1.3 billion technology and digital transformation fund. Zurich runs 160 AI use cases in production or advanced pilot. Combined, Europe’s four largest carriers are spending over EUR 12 billion on technology.

Every one of these deals describes the same thing: document processing, claims summarization, submission triage, customer service automation. The technology works. Zurich’s Sixfold AI integration saves approximately one hour per underwriting submission. Travelers achieves straight-through processing on more than 50% of eligible claims files. Liberty Mutual’s auto damage estimator is trained on 200 million data points across five million claims. These are real operational gains.

The gains are worth capturing. A carrier that saves 17,000 administrative hours per week or achieves straight-through processing on half its eligible claims is making a rational, defensible investment in expense-ratio reduction. Corporate officers are right to pursue these returns. But expense-ratio improvement through workflow automation is not a business model change. It is not new revenue. It is not a new product architecture. It is faster execution of an existing process through existing data. The distinction matters because the vendors, the consultants, and the PE-backed ventures are structured to sell optimization as transformation. The carrier gets a real operational gain. The vendor gets to claim an enterprise AI deployment. The PE sponsor gets a guaranteed return. Everyone profits. But the carrier’s competitive position relative to the sensor data opportunity scaling underneath the industry has not changed.

None of these deals address sensor data integration, edge inference, parametric product design, or the Physical AI data layer scaling underneath the industry. None move the carrier toward the new revenue sources that require new data. Every deal optimizes existing workflows through existing business models, faster.

As I documented in Optimized Satisficing earlier this year, carriers cycle through vendor partnerships while the organizational constraints remain unchanged. AIG’s public record illustrates the pattern across three technology eras in nine years: the Two Sigma venture (2016, divested), the Palantir partnership (2021, ongoing), the Anthropic relationship (2025, announced at Investor Day). The language was recycled nearly verbatim each time. This brief does not revisit that case. It examines who profits from the pattern and what it obscures.

The shift from FOMO to FOMU (Fear of Messing Up) is already visible. Early adopters found enterprise AI solutions had questionable data training practices, unpredictable results, and roadmap instability. BCG’s 2026 P&C Research Index found that only 38% of P&C insurers generate value at scale from AI within core workflows. AI spending as a share of carrier revenue is projected to triple in 2026, but measurable returns remain elusive. MIT’s Project NANDA reported that 95% of generative AI pilot programs failed to deliver measurable financial returns, though the research is early-stage and the sample reflects an adoption environment still in formation. Per S&P Global Market Intelligence, 42% of companies abandoned most generative AI initiatives in 2025, up from 17% the prior year. The NANDA findings are directionally consistent with what carriers report anecdotally, but should not be treated as a settled verdict on the technology’s potential.

Munich Re’s decision to shutter its $1.2 billion venture arm is a signal worth reading carefully. The 40-person team made approximately 100 investments over a decade. Munich Re shut it down despite recording EUR 2.1 billion in quarterly profit, bringing innovation under its asset management arm MEAG. When the most sophisticated corporate VC in insurance retreats from startup investing to concentrate innovation in core business, the signal is not that innovation failed. The signal is that the venture model failed.

The carriers that announced AI partnerships are not making bad technology decisions. The technology works. They are buying optimization of a business model that needs transformation, and the vendors selling to them are structured to profit from the optimization, not the transformation.

3. The Commodity Dynamic

The product these ventures deploy is commoditizing underneath them. But the commoditization is not producing the outcome most analysts expect. The model market is not consolidating toward a single winner. It is fragmenting, for structural reasons that the PE-backed consulting ventures cannot overcome.

The pricing collapse -- and the bifurcation underneath it. In early 2024, frontier language model inference cost roughly $60 per million tokens. By mid-2026, the commodity tier has crashed to $0.10 to $0.14 per million input tokens: DeepSeek V4 Flash at $0.14, GPT-4.1 Nano at $0.10, Gemini Flash-Lite at $0.10. The average output price decline since March 2023 is 94.5% using a price-index methodology benchmarked to GPT-4 at launch. Commodity tier output prices fell roughly 80% year over year, from approximately $0.50 to $0.70 per million tokens in mid-2025 to the $0.10 to $0.14 range by mid-2026. Enterprise token costs dropped from $18.40 to $6.07 in a single year, a 67% decline. Epoch AI documents a 50x annual decline rate in inference costs. For comparison, Moore’s Law delivered approximately a 2x improvement in transistor density every two years, or roughly 10x per decade. AI inference costs are declining at roughly 50x per year, a pace that ARK Invest and multiple analysts describe as outpacing Moore’s Law by 50 to 100 times. This is the fastest cost deflation curve in the history of computing.

But the pricing story has a second chapter that most analysis misses. While commodity prices crater, frontier pricing has held firm or increased with each new model release. GPT-5.5 runs at $5 input and $30 output per million tokens. GPT-5.5 Pro, the most expensive model on the market, costs $30 input and $180 output. Claude Opus 4.6 sits at $5 and $25. The spread between GPT-5.5 Pro at the frontier ceiling and GPT-4.1 Nano at the commodity floor is 450x on output tokens. Against the cheapest available model, DeepSeek V4 Flash at $0.28 per million output tokens, the spread widens to 643x. That is not a marginal pricing difference. It is a two-tier market.

The commodity tier is genuinely good news for carriers. A procurement team that routes 85% of queries through $0.10 models and reserves frontier capacity for the 15% that requires it can capture enormous value from the technology at a fraction of the cost. Enterprise routing benchmarks aggregated across multi-model proxy tools and infrastructure platforms indicate that 85% of production queries can be handled by budget-tier models, yielding 60-80% cost reduction through intelligent routing. The issue is not the model pricing. The issue is the consulting layer between the carrier and the model. PE-backed consulting ventures need carriers on the expensive tier. Multi-year frontier-tier capacity contracts at $5 to $30 per million tokens justify the need for expensive deployment engineers who can “optimize” the premium product. The 85% of queries that could run on budget models at a fraction of the cost are not a talking point in the consulting pitch.

OpenAI launched “Frontier” in February 2026, an enterprise platform offering 1-3 year guaranteed capacity contracts with tiered volume discounts. The pricing is contact-sales only. Multi-year compute commitments convert AI spend into something resembling cloud infrastructure contracts, which is exactly the procurement model insurance CIOs are familiar with and the consulting intermediaries are incentivized to recommend.

The switching reality. Enterprise model switching behavior requires careful parsing. Eighty-one percent of enterprises now use three or more model families. The average enterprise runs 4.2 AI models, up from 1.9 in 2023. Multi-model routing architectures cut costs 30% to 85% while maintaining 95% or higher quality. This looks like the model layer is a commodity. But actual annual vendor churn is only 11%, per Menlo Ventures survey data. Enterprises route API calls across multiple models aggressively but rarely switch the vendor relationship. The distinction matters: the 11% vendor churn rate explains why PE-backed consulting ventures can still extract revenue despite commodity pricing. The enterprise is locked into the consulting relationship even as it routes around the model. What changes primary model selection on any given day is latency, cost, or task fit. What changes the vendor relationship is a procurement cycle that moves at the speed of enterprise bureaucracy, not the speed of API routing.

Per Menlo Ventures enterprise tracking data, OpenAI’s enterprise spend share has fallen from approximately 50% to 27%. Anthropic’s share rose to 40% by December 2025, driven by Claude Sonnet 3.5’s coding performance, then began facing the same competitive pressure. The duopoly is collapsing in terms of model usage even as the vendor relationships persist. Per the AICC 2026 AI API Infrastructure Report, open-source and open-weight models captured 38% of enterprise token volume in Q1 2026, up from 11% in Q1 2025, based on anonymized data across more than 2.4 billion API calls.

Why the market fragments rather than consolidates. The database market provides the structural analogy. When distributed databases emerged, engineers believed a single database would serve all workloads. The CAP Theorem proved otherwise: no distributed system can simultaneously guarantee consistency, availability, and partition tolerance. The market fragmented along workload lines into relational, document, graph, time-series, columnar, and key-value stores. No single database won because the constraints were mathematically irreconcilable.

The LLM market faces the same structural logic. No single model simultaneously optimizes for speed, cost, accuracy, privacy, domain specialization, and latency. Different use cases optimize different constraints. Insurance underwriting prioritizes accuracy and regulatory compliance. Customer service prioritizes speed and cost. Coding prioritizes context window and instruction following. Edge deployment prioritizes size and latency. The market fragments not because the models are bad but because the workloads are heterogeneous.

The Herfindahl-Hirschman Index for the enterprise LLM market suggests a moderately concentrated oligopoly, not a winner-take-all structure. This is not the market structure that justifies a $10 billion consulting venture built around a single model provider.

The specialist unbundling. Domain specialists are capturing revenue that once flowed to generalist model providers. The pattern is consistent across categories, and the Corpus Search evidence maps it in detail.

In customer service, Intercom’s Fin Apex 1.0 achieves a 73.1% resolution rate, outperforming GPT-5.4 and Claude Opus 4.5. Intercom built a proprietary post-trained model and charges $0.99 per resolved conversation. Anthropic and OpenAI receive zero revenue. Decagon routes 80% of model traffic through its own trained models. Sierra AI runs 15 or more models simultaneously across 40% of Fortune 50 companies and plans to replace third-party models entirely with custom-trained alternatives.

In document processing, Reducto’s vision-first parsing scores approximately 0.90 on RD-TableBench versus 0.81 for Google Document AI, using proprietary models. Rossum’s Aurora processes documents for 450 organizations including Bosch and Siemens on a proprietary transactional LLM. Generalist LLMs treat document pages as flat, serialized text, destroying the spatial and tabular relationships that document processing requires. The specialists solved this with vision-first pipelines that bypass LLM APIs entirely.

In software development, the unbundling is happening in real time. Windsurf (Codeium) reports processing context at 950 tokens per second through its proprietary SWE-1.5 model served on Cerebras hardware, approximately thirteen times faster than Claude Sonnet 4.5. Cursor, once a distribution channel sending API revenue to Anthropic, shifted its backend to Kimi K2.5 from Moonshot AI, which scores 76.8% on SWE-Bench Verified versus Claude Sonnet 4.6 at approximately 72%. Today’s distribution channels are tomorrow’s revenue competitors.

In insurance specifically, the specialist ecosystem runs deeper than most coverage acknowledges. Shift Technology is trusted by four of the top five US P&C insurers, operating natively inside Guidewire ClaimCenter with a hybrid engine combining statistical data science with generative AI agents. Sixfold AI runs quantized open-weights models within isolated single-tenant environments on carrier Azure budgets. Indico Data operates 80 proprietary models across 120 product lines. Tractable’s computer vision for damage assessment is not LLM-based at all. Earnix deploys predictive and agentic AI directly into the underwriting, pricing, and product personalization cycle, allowing actuaries to adjust and execute pricing rules in real time. For actuarial work, ActuBench found that a Gemma 4 model on consumer hardware sits on the cost-performance Pareto front, within one item of the leaderboard top.

In compliance and regulatory technology, the unbundling is already complete. The RegTech AI market, valued at $14.9 billion and projected to reach $107 billion by 2035, is 95% independent of generalist LLMs because it predates them entirely. Financial institutions are deploying “dual-model architectures” where a primary generative model produces outputs and secondary specialized models audit those outputs in real time for bias, numerical errors, and regulatory violations, all hosted within isolated virtual private clouds where no client data touches external servers.

The trajectory follows what I call the “specialist start, proprietary end” pattern. Companies start on Claude or GPT APIs, accumulate domain-specific data, train proprietary models that outperform generalists on their specific tasks, and progressively reduce API dependency. Based on Indenseo’s company-level analysis across eight specialist categories, approximately 35% of specialist AI revenue still functions as a distribution channel for generalist LLM providers while an estimated 65% has become competitive, using proprietary or open-source models. The split is trending toward competitor. Two categories are already fully unbundled: structured data extraction (where constrained decoding libraries mathematically guarantee output compliance, eliminating the need for generalist API calls) and visual damage assessment (where proprietary computer vision has no LLM dependency). Four categories are in active transition. Only enterprise conversational AI remains structurally dependent on frontier reasoning models.

The financial exposure. OpenAI’s revenue is growing: CFO Sarah Friar confirmed the company surpassed $20 billion in annualized revenue in 2025, tripling from approximately $6 billion in 2024, with $29.4 billion projected for 2026. But the company’s cost structure remains inverted. Compute and technical talent consume approximately 75% of total revenue. Gross margin sits at approximately 33% against a SaaS industry benchmark of 75% to 85%. HSBC estimates OpenAI requires $207 billion or more in additional capital through 2030. That figure exceeds the GDP of New Zealand and approaches the GDP of Portugal: a single company’s capital needs rival the annual economic output of a developed nation. The company has $1.4 trillion in long-term compute commitments, a figure approaching the GDP of Australia ($1.7 trillion) or Spain ($1.6 trillion): a single company has committed to spending on computing infrastructure what entire developed nations produce in a year. OpenAI revised this target down to approximately $600 billion by 2030 in February 2026, meaning even the company’s own infrastructure projections are contracting. Twelve or more senior executives departed in 2025, with a triple exit in April 2026. Revenue is growing. The economics underneath the revenue are not.

Anthropic’s trajectory is different in degree but not in kind. The company reached a $30 billion annualized run rate by April 2026, with 80x revenue growth in Q1 2026. Claude Code alone generates $2.5 billion in annualized revenue. Anthropic holds 40% of enterprise LLM spend, ahead of OpenAI at 27%. Its burn rate is projected to drop to approximately 9% of revenue by 2027, compared to OpenAI’s 57%. These are real strengths. But 75% of that revenue flows through platforms controlled by Amazon, Google, and third-party tools. Amazon is building Nova models that directly compete with Claude on the same Bedrock platform where Anthropic generates most of its revenue. When your largest distribution partner is also your most motivated competitor, the revenue looks less durable than the growth rate implies.

Both companies launched PE-backed consulting ventures within 36 hours. That timing is not coincidental. It is a coordinated market signal that neither company’s model revenue alone can sustain the growth trajectory their valuations require.

The production reality. Vera Engineering documented removing all OpenAI from its production platform, citing poor scaling, high latency, lack of version control, and neglect of mid-size enterprises. A former OpenAI engineer described the internal codebase after growing from 1,000 to 3,000 employees in a single year. Independent testing documented measurable API degradation in Anthropic’s long-context coherence. RAG systems, the backbone of most enterprise AI deployments, fail silently in production through retrieval drift, embedding decay, and the “lost in the middle” problem where models lose track of information in long contexts. Indian venture capital data provides a retention signal: AI copilots claiming 30% time savings showed 15% to 25% ninety-day retention. Meanwhile, “boring” compliance automation with 80% retention raised capital in 45 days. The tool that sounds transformative does not stick. The tool that automates a specific, measurable workflow does.

The clearest framing came from conversations with sensor OEM representatives who view models as “a feature, not a product.” The observation lands differently when you consider that telematics companies have been running machine learning models for more than twenty years. Nobody calls Geotab or Samsara an ML company. Samsara, with $1.9 billion in annual recurring revenue and 55 million connected sensors, describes itself as a “Connected Operations Platform.” Geotab, with more than five million connected vehicles, calls itself a “connected transportation solutions” provider. Karooooo, with 2.6 million subscribers across Africa, uses an “intelligent SaaS platform.” The ML is embedded. It is invisible. It is infrastructure. That is the trajectory. The PE-backed ventures are building a consulting business around a product category following the same arc.

4. Three Paths

Three distinct purchasing paths exist for carriers buying AI capabilities. They are not equivalent, and the differences matter more than most carrier procurement processes acknowledge.

Path 1 is the direct model company deal. The carrier signs with Anthropic or OpenAI, gets document processing and claims summarization, and receives FDEs at $200,000 to $300,000 per year. The carrier pays for a commodity product plus expensive humans, with PE taking a guaranteed return. The carrier gets optimization of existing workflows. It does not get access to new data sources, new product architectures, or the sensor layer where insurance-relevant data is being generated. Path 1 is also a distribution subsidy: the carrier’s contract helps the model company demonstrate enterprise revenue to justify its valuation, while the PE sponsors extract their guaranteed 17.5% regardless of whether the carrier sees measurable operational improvement. And the consulting intermediary steers the carrier toward multi-year frontier-tier capacity contracts at $5 to $30 per million tokens when 85% of production workloads could run on models priced at $0.10 to $0.50.

Path 2 is the platform deal. Microsoft Azure AI Foundry now offers access to more than 11,000 models, including Claude, DeepSeek, Mistral, Llama, and Microsoft’s own MAI family launched in April 2026. The value proposition is model optionality with enterprise security. Interviews with hyperscaler representatives confirm the strategy: the platform is a foundry, not an exclusive relationship. The carrier can switch models as the market evolves without renegotiating vendor contracts. Given that 81% of enterprises already use three or more model families and multi-model routing cuts costs 30% to 85% while maintaining quality, the foundry approach aligns with how enterprises actually consume AI. Path 2 is rational cloud infrastructure. But it is still a cloud play. It does not move the carrier toward sensor data.

Path 3 is edge and Physical AI. On-device inference. Sensor data processed at the point of collection. No cloud dependency for time-critical decisions. The parametric insurance opportunity. The predictive maintenance opportunity. The mechanical health usage-based insurance opportunity. The ESG monitoring opportunity. Path 3 is where the new products are.

The structural point: these paths are not a progression. A carrier on Path 1 does not graduate to Path 3 by buying more LLM tokens. The data sources are different. The engineering requirements are different. The vendor relationships are different. The entire technology architecture is different. Path 1 does not physically prevent a carrier from pursuing Path 3. A multi-billion-dollar insurer can, in theory, parallel-track both. But in practice, Path 1 consumes the scarce resources that Path 3 requires: not just budget, but engineering leadership bandwidth, board-level attention, and the organizational willingness to absorb a second major technology initiative while the first is still being justified. The carriers that have signed Path 1 deals are not unable to pursue Physical AI. They are unlikely to, because every quarterly review cycle spent demonstrating ROI on the LLM engagement is a cycle not spent building sensor data capabilities that have no existing vendor relationship to anchor them.

The local inference trajectory reinforces the point. At Google I/O 2026, Google announced Gemini Intelligence embedded in Android 17 for multi-app task automation on flagship devices, while also expanding web-based Gemini automation features to phones with 4GB or more of RAM. Apple’s WWDC 2025 revealed a rebuilt Siri powered by on-device models with a 3-billion-parameter foundation model available to all developers through a Swift API at zero cost. Google distributes AI through 3 billion Android devices and Chrome’s 65% browser share. Apple distributes through 2.5 billion active devices. Both companies give inference away because they make money from advertising, hardware, and ecosystem, not from inference. The carrier’s Path 1 vendor sells inference. The platforms that control the surfaces enterprises interact with are making inference free. That is a structural mismatch the consulting venture cannot resolve.

Cloud-dependent AI also faces a growing infrastructure wall that local inference sidesteps entirely. Data compiled by the Lawrence Berkeley National Laboratory (LBNL) shows US data centers consumed roughly 176 terawatt-hours of electricity in 2023, accounting for 4.4% of total US consumption, and are projected to reach 325 to 580 terawatt-hours by 2028. Data centers now account for half of all new US electricity demand growth. Water consumption hit 17.4 billion gallons in 2023, projected to reach 38 to 73 billion gallons by 2028. In regulatory filings, NV Energy notified regulators it is redirecting 75% of the electricity supply serving 49,000 Lake Tahoe-area residents to data centers in Northern Nevada. Public records requests in Fayette County, Georgia revealed that a QTS data center facility secretly consumed 29 million gallons of unmetered water over 15 months before residents detected it through low water pressure. The North American Electric Reliability Corporation (NERC) issued a Level 3 alert, its highest operational severity, over data center load disruptions threatening grid reliability. PJM Interconnection, which serves 65 million people across 13 states, projects a six-gigawatt capacity shortfall by 2027: roughly the output of six large nuclear reactors, or enough electricity to power approximately 4.8 million homes. National project trackers confirm that more than $64 billion in data center projects have been blocked or delayed by community opposition in the past two years. At least 142 activist groups across 24 states are organizing against construction, and 14 states have enacted temporary construction pauses. These constraints do not mean a carrier’s API queries will be physically throttled tomorrow. They mean the operating cost trajectory for cloud-dependent AI infrastructure is rising while edge inference costs continue to fall. Edge inference requires no new construction, no water cooling, no grid interconnection queues, and no zoning approvals. It runs on hardware the customer already owns.

The hidden cost curve of integration middleware reinforces the lock-in. Enterprise middleware follows what one analysis calls a “vicious cost cycle”: affordable at launch, cost-unpredictable at scale. The total cost of ownership gap between projected and actual spend grew to nearly 40% at one financial services firm. Organizations three to five years into a middleware relationship have hundreds of integrations, trained teams, and interconnected APIs. Switching feels impossible. That is the structural dependency the PE-backed consulting ventures are designed to create.

A Sri Lankan architect who transformed one of Asia’s oldest insurers offers the counter-model. He built integration layers over legacy systems rather than pursuing the rip-and-replace approach Western consultancies recommend. His track record: 40% to 50% premium growth versus competitors stuck in multi-year platform migrations. His advice to carriers in emerging markets: stop listening to big consultancies. His rule: AI projects that do not connect directly to revenue metrics do not survive budget cycles.

5. The Physical AI Blindspot

The insurance industry has no seat at the Physical AI table. This is not an opinion. It is a quantifiable absence.

Edge Impulse, the company Qualcomm is acquiring along with its 170,000-developer community, has published more than 18 case studies. Only two mention insurance, and neither originated from Edge Impulse itself. They were identified through our analysis, not through the company’s own market targeting. Syntiant, which has shipped more than 50 million edge AI devices, has zero insurance mentions in its published materials. Bosch Sensortec, with 300 million devices running Qeexo AutoML, does not list insurance as a target market. The companies building the most insurance-relevant technology do not see insurance as a market.

The edge AI chip market is scaling from $9.5 billion in 2025 to a projected $57.8 billion by 2034. For scale, $57.8 billion would exceed the current global cyber insurance market ($16.7 billion in 2025) by more than three times, or roughly match the global property catastrophe reinsurance market. The edge AI chip market is on track to become larger than many insurance verticals that carriers consider core business. Small language models deliver local inference in 50 to 200 milliseconds versus 2 to 5 seconds for cloud-hosted LLMs. Neural processing unit efficiency exceeds 200 TOPS per watt in 2026, up from under 40 four years prior. Half-billion-parameter models now run inference on a Raspberry Pi in two to four seconds. Apple’s depth-split architecture applies a 5:3 layer-sharing ratio within its local model blocks to achieve a 37.5% reduction in KV cache memory usage, while separate flash-memory mapping techniques enable devices to run models that exceed their raw RAM capacity. Apple’s Private Cloud Compute, built on custom silicon with Secure Boot and Secure Enclave, contains no privileged interfaces, no remote shells, and no debug tools, making it architecturally impossible for any operator, including Apple, to access user data in the cloud compute layer.

The telematics precedent is instructive. Insurance did not pick the winners in fleet telematics. The technology was built for fleet management, logistics, and auto OEMs. Insurance consumed what existed, often poorly. As I documented in Why Insurance Telematics Integrations Fail for Carrier Management, the same five failure modes appear regardless of carrier size, vendor selection, or technical architecture: the legacy architecture trap, data quality mismatch, workflow adoption failure, timeline mismatch, and departmental silos. Physical AI is the same pattern repeating at a deeper hardware layer.

What makes the current moment different is what is happening in markets that most Western carriers are not watching.

In Kenya, the Kilimo Salama program (Syngenta Foundation, UAP Insurance, Safaricom) operates parametric crop insurance on a sensor-to-payment stack: automated solar-powered weather stations broadcast rainfall data over Safaricom’s 3G network. When rainfall drops below a regional threshold, payouts trigger directly via M-Pesa. No claims forms. No adjuster. No chatbot. No LLM. The program started with 200 farmers in 2009 and has scaled to more than 1.7 million smallholder farmers across five countries, with $181 million in total sum insured.

ACRE Africa’s picture-based insurance combines satellite imagery with smartphone photographs of crops at growth stages, registration via USSD, and premiums and payouts via M-Pesa. In Rwanda, Sawa Telematics processes more than 10 million data points per day across over 3,000 connected vehicles at approximately $9 per vehicle per month. In South Africa, MTN and Huawei deployed OBD telematics devices feeding data directly to insurers for usage-based pricing. In India, Reliance launched JioMotive, an OBD-based telematics system retailing at approximately $79, targeting vehicle segments that factory-installed telematics does not reach. ICICI Lombard launched pay-as-you-use and pay-how-you-use motor insurance using telematics built independently of Western consulting firms.

The distribution infrastructure is already in place, with USSD codes effectively serving as the operating system for African insurance delivery. This model thrives through both dedicated insurance channels, like Sanlam Nigeria (1056#) and Heirs Insurance (1100#), and embedded micro-insurance products integrated directly into everyday mobile network menus, such as Safaricom Kenya (544#), Airtel Kenya (334#), and Telkom Kenya (444#). Monthly premiums start at $0.24. Insurance is bundled with mobile data plans. The Sanlam/MTN aYo alliance distributes micro-life and hospital-cash insurance across 19 African countries, with 6.3 million active policies and a target of 30 million policyholders leveraging MTN’s 100 million active mobile wallets.

These implementations share a structural feature: the sensor-to-payment stack (weather stations, satellite, telematics, IoT sensors to USSD/SMS to M-Pesa and mobile money) has no natural entry point for the LLM consulting layer. The $200,000-per-year Forward Deployed Engineer has no role in a system where a solar-powered weather station triggers an M-Pesa payment.

A necessary caveat: Western commercial insurance operates under fundamentally different regulatory frameworks, risk pool structures, and capital requirements than Kenyan parametric crop insurance or Rwandan fleet telematics. The emerging markets evidence is included as architectural proof-of-concept for sensor-to-payment loops, not as a direct operational blueprint for a US commercial carrier. What it demonstrates is a principle: insurance products can be built on sensor-to-payment stacks without an LLM consulting layer in the value chain. The principle scales even where the specific implementation does not. The question for Western carriers is not whether to replicate M-Pesa-based parametric payouts, but whether the billions flowing into LLM optimization are building any capability that connects to the sensor data architectures where the next generation of insurance products will be designed.

Per Britam’s published corporate reporting, Britam Kenya built its AI Motor Assessment Service internally at BetaLab: image authenticity validation, damage severity detection, real-time repair cost estimation, settlement via M-Pesa, claims resolved within two hours. Not built by a $10 billion PE-backed consultancy. Built by a Kenyan insurer’s internal innovation lab. The Association of Kenya Insurers partnered with Microsoft Azure (not a consulting venture) for industry digitization, reporting 30% management cost reduction and 40% claims expense reduction per a published Microsoft partnership case study. Cloud platform play. Not consulting engagement.

The LoRaWAN soil moisture sensor market, valued at $1.8 billion in 2025, is projected to reach $4.9 billion by 2034. Agri-insurance firms are incorporating verified soil moisture monitoring into crop insurance underwriting. Smart contract parametric insurance from ACRE Africa and Etherisc delivered blockchain-based weather index coverage to more than 12,900 Kenyan smallholder farmers at an estimated 30% to 40% cost reduction versus traditional products.

The carriers spending billions on Path 1 AI deals are buying optimization of existing Western insurance workflows. The carriers they are not watching are building insurance products on sensor-to-payment architectures that those AI deals cannot serve.

6. The PE Playbook

Private equity firms have a business model. They standardize back-end operational functions across portfolio companies, drive efficiency through consistency, and extract returns through a proven playbook. It is a very successful model. It has generated extraordinary returns for decades. It is also a model designed for PE economics.

The logic extends to technology vendor selection. Tech startups tend to favor Google Workspace. PE firms often standardize portfolio companies on Microsoft Office, in part because it simplifies back-office integration when a portfolio company is acquired by a larger enterprise. The choice is rational for PE economics: consistency across a portfolio of 200 companies reduces operational friction at every stage from acquisition through exit. It would not necessarily be rational for a company making the decision purely on its own operational merits, where the best tool for its specific workflows might be different. The same logic applies to standardizing portfolio companies on a specific LLM vendor. When PE firms want exposure to novel, non-standard technology, they invest through venture arms built for that purpose. DeployCo is not a venture arm. It is the PE operational playbook applied to AI: standardize every portfolio company on a specific model vendor, whether or not that vendor relationship is the best solution for any individual company’s needs. Carriers should not mistake that playbook for a technology strategy.

The 17.5% guaranteed annual return to DeployCo’s PE sponsors over five years is the PE model working exactly as designed. It is a return structure, not a technology outcome metric. A technology investment generates returns when the technology produces value for the customer. A guaranteed-return structure generates returns when the customer pays. When the customer base is captive (2,000+ portfolio companies controlled by the PE sponsors themselves), the risk of non-payment approaches zero. Vista Equity takes the standardization logic further. As reported by the Financial Times in November 2025, Vista formally told its LPs it would score every portfolio company on AI adoption velocity and tie those scores to capital allocation decisions across its 90-plus software companies. This is operational standardization at scale, consistent with the PE playbook, applied to AI vendor selection.

The revenue mechanics are straightforward. FDE job postings confirm base compensation of $150,000 to $250,000 plus equity, with total compensation in the $200,000 to $300,000 range. The billing rate to the enterprise client is a multiple of that. Consulting firms that took equity positions in DeployCo bill senior consultants at $250 to $500 or more per hour for AI-specific work. The model itself is approaching commodity pricing. The humans required to make the model work are not. And the pricing bifurcation in the model market serves the consulting economics: steering clients toward frontier-tier contracts at $5 to $30 per million tokens rather than commodity alternatives at $0.10 to $0.50 justifies the need for expensive deployment engineers who can “optimize” the premium product. The 85% of queries that could run on budget models at a fraction of the cost are not a talking point in the consulting pitch.

The labor economics at the foundation of this system are worth examining. A TIME investigation documented that OpenAI paid its data labeling contractor Sama approximately $12.50 per hour. Sama paid the workers who built the safety systems underlying ChatGPT $1.32 to $2.00 per hour, an intermediary capture rate of 84% to 89%. Andela, a PE-backed engineering talent marketplace valued at $1.5 billion, sources software engineers from African markets at a fraction of the cost of US-based FDEs. The margin structure illuminates the incentive. When the model is a commodity and the consulting is the revenue, the incentive is to maintain complexity, not resolve it.

The FDE model creates structural dependency: the enterprise needs the FDE because it cannot do the work itself. The CIO functions as a purchasing agent, not an engineering leader, selecting vendors rather than building capability. The FDE engagement stretches because knowledge capture is irreducible, and the enterprise lacks the talent to assess whether the engagement should end. The FDE leaves; the enterprise needs another one. FDE demand has surged 800% since January 2025. Rippling credits FDEs with 90% customer retention at $500 million in ARR. These numbers demonstrate that the model works as a business. They do not demonstrate that it works as a technology outcome for the buyer.

The PE playbook is rational on its own terms. The standardized AI engagement gives each portfolio company an “AI transformation” narrative that can support exit multiples. The 17.5% guaranteed return on DeployCo capital is a protected income stream, structurally isolated from any individual portfolio company’s operational performance. The PE firm collects the consulting return with certainty while the narrative value of “AI-enabled operations” supports the exit thesis. This is sound PE economics. The question is whether a carrier, which is not a PE portfolio company and does not exit on PE timelines, should adopt a technology strategy designed for portfolio-wide standardization rather than for its own specific operational and competitive needs.

The distribution dynamics make the economics more fragile than they appear. The PE ventures create distribution by routing demand through captive portfolio companies and consulting relationships. But this is manufactured distribution, not structural distribution. Google, Microsoft, Apple, and Amazon distribute AI through billions of devices, operating systems, browsers, and cloud platforms. They give inference away because they monetize the ecosystem. DeployCo distributes AI through 2,000 portfolio companies. The scale mismatch is not close. And the 11% annual vendor churn rate that makes PE ventures viable today depends on enterprise procurement inertia, not product superiority. As multi-model routing becomes standard and the specialist unbundling continues, the consulting relationship becomes the only durable revenue source. The model revenue is already commodity. The consulting fee is next.

As model pricing collapses and open-source alternatives close the performance gap, deployment yield compression accelerates: each dollar of consulting spend produces less measurable enterprise value than the last, but the engagement structure is designed to persist regardless.

TPG’s capital allocation tells a more complete story than any press release. TPG led the $10 billion DeployCo investment. TPG’s Rise Funds also led a $350 million investment in Cambridge Mobile Telematics, the smartphone-based telematics company, with co-investors Allianz X and State Farm. TPG is simultaneously the largest investor in the LLM consulting venture and a leading investor in the Physical AI data layer. The distinction is revealing. When TPG wants standardized AI deployment across portfolio companies, it uses DeployCo. When it wants exposure to novel technology that does not fit the standardization playbook, it uses Rise. PE firms are disciplined about which instrument serves which purpose. Carriers buying into these ventures would benefit from applying the same discipline to their own technology decisions.

NVIDIA’s counter-bet deserves attention. Jensen Huang open-sourced NemoClaw, a security framework for AI coding agents that adds safety controls and governance guardrails to autonomous AI development workflows. The bet is explicit: developer competence, secured by open tooling, scales better than consulting armies. If enterprises can deploy and govern AI agents without $200,000-per-year humans, the PE-backed consulting venture loses its revenue engine. The model commoditizes. The consulting commoditizes. What remains is the data, the sensor infrastructure, and the engineering talent to connect them.

7. The Execution Question

Recognizing the asset class mismatch is necessary but not sufficient. The execution challenges facing any carrier that wants to move from Path 1 to Path 3 are real, and pretending otherwise would be dishonest.

The purchasing-agent CIO cannot evaluate whether the FDE deployment worked because the organization lacks the engineering talent to assess outcomes. The AI-insurance talent intersection is, by multiple accounts, “exceptionally narrow.” Carriers posting Silicon Valley skill stacks at insurance compensation levels are not solving this. The 400,000 insurance professionals projected to exit the industry are taking institutional judgment with them, and the AI systems being deployed without capturing that expert knowledge will produce average-performing outputs at best.

The brain drain is cognitive, not mechanical. Experienced underwriters report identical language across carriers: “I’m not burned out from underwriting. I’m burned out from everything around it.” Senior underwriters spend meaningful time reviewing out-of-appetite submissions, re-entering data across systems, and writing extended documentation for modest pricing decisions. They start trading clarity for efficiency, declining good-but-complex risks to avoid downstream friction. The result is portfolio opacity risk: invisible drift in book composition that no AI summarization tool will detect because the decision data it would need was never captured.

The carrier’s Path 1 deal crowds out budget and attention for everything else. Forrester projects that total US technology spending will increase by $173 billion in 2026, growing 7.8% year over year. Insurance represents approximately 6% of total US technology spending, placing carrier technology budgets in the range of $130 billion to $145 billion annually. For scale, that exceeds the combined annual revenue of Travelers ($43 billion), Hartford ($25 billion), and CNA ($14 billion). Insurance as a vertical spends more on technology than most readers would expect. More than half of AI budgets are directed at sales and marketing tools while back-office automation (the highest-ROI application) is neglected. The board sees “AI” as checked. The actual sensor data opportunity requires engineering talent, actuarial framework innovation, and OEM relationships that no current carrier deal provides. At one major technology company, a team burned through its entire 2026 AI coding budget in four months; individual engineer token costs ran $500 to $2,000 per month. The CTO sent staff “back to the drawing board.” Budget overruns on LLM consumption are not hypothetical. They are happening now, and they crowd out investment in everything else.

The AI coding arms race provides a preview of what happens when organizations deploy powerful tools without the governance to manage them. Stanford’s Digital Economy Lab documented a 13% decline in entry-level hiring within AI-exposed occupations, concentrated among 22-to-25-year-olds. The deskilling pattern has specific operational symptoms that industry accounts describe consistently: pull request volume outpacing comprehension as junior engineers generate code through AI assistants without fully understanding the output, incentive structures rewarding output quantity over code quality, and collaborative communication declining as engineers interact with AI tools rather than with each other. Senior engineers at multiple firms describe an approaching “AI maintenance wall” at year two to four when AI-generated systems become unmaintainable. These accounts are illustrative of a broader pattern, not systematic evidence on their own. But the pattern they illustrate, organizations deploying powerful tools faster than they can build the governance to manage them, is consistent with the documented budget overruns, the Stanford hiring shifts, and the 42% initiative abandonment rate documented by S&P Global Market Intelligence. The deskilling risk is not theoretical.

The opportunity for operators who bridge Physical AI hardware to insurance operations is substantial: parametric products, mechanical health usage-based insurance, sensor-driven underwriting, real-time risk management. The edge AI chip market alone is scaling toward $57.8 billion. But the path requires engineering competence that the current purchasing pattern actively erodes by consuming the budget, the executive attention, and the organizational bandwidth that sensor integration would demand.

At what point does satisficing become a fiduciary question? The asset class mismatch sharpens that question. Carriers are not merely accepting “good enough” performance from their technology investments. They are committing billions to vendors whose models are commoditizing, whose specialist competitors are training proprietary alternatives that outperform generalists on insurance tasks, whose revenue depends on platforms controlled by competitors building cheaper alternatives, whose entire paradigm cannot address the sensor data opportunity scaling underneath the industry, and whose cloud infrastructure faces growing political, regulatory, and environmental resistance that edge and Physical AI architectures do not carry.

The regulatory environment is accelerating the shift. The EU AI Act imposes fines up to EUR 35 million or 7% of global revenue for prohibited AI practices. The EU Data Act exempts data processed on-device and deleted from certain compliance requirements, creating a structural regulatory incentive for edge processing. Zero-egress mandates in South Korea, China, and European sectors legally disqualify sending sensitive data to cloud AI. The regulatory trajectory favors on-device processing, which favors the Physical AI stack, which favors the carriers that invested in sensor infrastructure over the carriers that invested in LLM consulting contracts.

The carriers writing checks to PE-backed AI consulting ventures are not making technology decisions. They are making capital allocation decisions. And the question is whether those decisions serve the carrier’s own policyholders and shareholders, or whether they serve the private equity sponsors who structured the guaranteed return before the first engineer was deployed.

Companion Tool: Carrier AI Deal Assessment Scorecard

This intelligence brief includes a downloadable diagnostic scorecard designed for carrier strategy sessions. Fourteen indicators across three categories: capital structure exposure, model optionality, and Physical AI readiness. Score your current AI vendor relationship on a 0-to-2 scale to identify whether your deal addresses sensor data, model flexibility, and edge infrastructure, or whether it is a FOMO purchase. The scorecard is available to paid subscribers as a PDF download below.

Download Scorecard

This analysis is based on primary research, direct industry conversations, and two decades of operational experience in telematics, IoT data markets and insurance technology integration. Statistics are cited to primary sources where available.

Kevin Henderson is the founder and CEO of Indenseo and host of the Structural Signal podcast.

Notes

[1] The 35/65 specialist split (approximately 35% of specialist AI revenue functioning as a distribution channel for generalist LLM providers, 65% competitive using proprietary or open-source models) is an Indenseo market estimate based on company-level analysis across eight specialist categories. This is original research, not a published third-party metric.

[2] The observation that models are “a feature, not a product” is attributed to conversations with sensor OEM representatives. Primary interview intelligence, not attributable by name.

[3] Azure AI Foundry model-agnostic strategy and the characterization of the platform as “a foundry, not an exclusive relationship” are attributed to interviews with hyperscaler representatives. Primary interview intelligence, not attributable by name.

[4] The Sri Lankan architect case study (Section 4) is drawn from a published industry interview. The individual is not named to preserve the confidentiality of the original source conversation.

[5] Industry accounts describing the “AI maintenance wall,” deskilling patterns, and underwriter brain drain narratives (Sections 6 and 7) are compiled from multiple anonymized conversations with practitioners at US carriers and technology firms. These accounts are presented as illustrative of a broader pattern, not as systematic evidence.

[6] The “AI-insurance talent intersection is, by multiple accounts, ‘exceptionally narrow’” characterization (Section 7) reflects consistent findings across Indenseo’s primary research interviews, the Talent-Technology Intersection research report, and published industry workforce analyses.

DISCLAIMER AND LEGAL NOTICES

Important Disclosures

This report is published by Indenseo and distributed through Structural Signal Intelligence. It is intended for informational and educational purposes only and does not constitute investment advice, financial advice, legal advice, or a recommendation to buy, sell, or hold any security, investment product, or financial instrument. Nothing in this report should be construed as an offer or solicitation to make any investment.

The analysis, opinions, and conclusions in this report reflect the views of the author as of the date of publication and are subject to change without notice. While every effort has been made to ensure the accuracy and completeness of the information presented, Indenseo makes no representation or warranty, express or implied, regarding the accuracy, reliability, or completeness of any information contained herein. Data from third-party sources is believed to be reliable but has not been independently verified in all cases.

Readers should conduct their own due diligence and consult with qualified financial, legal, and tax advisors before making any investment or business decision based on information in this report. Past performance of any company, market, or investment discussed in this report is not indicative of future results.

The author and Indenseo may have direct or indirect relationships, advisory roles, or financial interests involving companies, executives, or markets discussed in this report. These relationships do not influence the analytical conclusions, which are based solely on publicly available information, primary source research, and the author’s independent professional judgment.

Copyright Notice

Copyright 2026 Indenseo Corporation. All rights reserved.

This report and its contents, including all text, data, analysis, frameworks, and original research, are the intellectual property of Indenseo. The frameworks, terminology, and analytical constructs introduced in this report, including the LLM FOMO Trap, the three carrier AI purchasing paths framework, the sensor-to-payment stack concept, and the capital-structure confession framing, are original work product of Indenseo.

Terms of Use

This report is licensed for use by the individual or organization that purchased it. The following terms govern its use.

Permitted uses: Subscribers may read, store, and reference this report for internal business purposes. Subscribers may cite brief excerpts (not to exceed 200 words per excerpt) in internal memoranda, presentations, or reports, provided that each citation includes the attribution: “Source: Structural Signal Intelligence, Indenseo.” Subscribers may reference the report’s analytical conclusions and frameworks in internal discussions and decision-making.

Prohibited uses: This report may not be redistributed, forwarded, published, posted online, or otherwise shared with any individual or organization that has not purchased a subscription, whether in whole or in part, without the prior written consent of Indenseo. This report may not be reproduced, photocopied, or stored in any retrieval system accessible to non-subscribers. The report’s original frameworks, models, and analytical constructs may not be presented as the work of any party other than Indenseo. This report may not be used in any public filing, prospectus, offering memorandum, or regulatory submission without a separate licensing agreement.

A commercial auto underwriting team at a top-twenty North American carrier has spent two years integrating telematics data from a TSP data aggregator into its pricing and risk selection models. The aggreg

ator occupies a critical position in the carrier’s data architecture: it normalizes data from dozens of fleet telematics service providers, each with its own hardware, data formats, and transmission protocols, into a unified feed. That feed delivers two layers of value. The first is anonymized, benchmarking-grade data across thousands of fleets that the underwriting team uses to calibrate risk models against industry-wide driving patterns. The second is policyholder-specific data from fleets that are both TSP customers and the carrier’s insureds, enabling individualized risk scoring that has improved the book’s loss ratio by four points over eighteen months.

Then the aggregator exits the business.

Read more

It starts at the Strait of Hormuz. 138 vessels a day crossed the strait before the February strikes. By March 9, one ship crossed. The waterway was open. A cancellation notice from an insurance company in Norway stopped 20 percent of the world’s oil supply. Brent crude rose from $71 to $103 in two weeks. The U.S. government launched a $20 billion reinsurance facility. JPMorgan estimated the actual need at $352 billion. Zero ships transited under the program.

Then the cascade reached helium. 200 containers stranded. A 45-day boiloff clock. The U.S. strategic reserve sold for $460 million. $650 billion in AI infrastructure now exposed. The SK Hynix chairman warned publicly of a potential wafer shortage exceeding 20 percent lasting until 2030.

The episode also covers California (668,000 homes on the FAIR Plan; $724 billion in exposure; an actuarial need of 80 percent), Australia (premiums up 178 percent in a decade), and Florida (the case study that shows what a functioning operating layer looks like when litigation noise is removed).

It ends with the dual failure: supply chain disruption from one direction and AI coverage withdrawal from the other, happening at the same time, to the same economic activity.

Both companion intelligence briefs are linked below.

Companion intelligence brief:

Companion intelligence brief:

Insurance:

Structural Signal

Why most data monetization projects fail — and what the survivors do differently.

By Kevin Henderson

LinkedIn: https://www.linkedin.com/in/kevinhenderson

©️2026 Indenseo Corporation

On March 24, 2026, Structural Signal published When the Operating Layer Breaks, analyzing how insurance cancellations at the Strait of Hormuz shut down 20% of the world’s oil supply in 72 hours and triggered cascading disruptions through LNG, fertilizer, petrochemicals, and global energy markets. That analysis traced the downstream cascade through the commodities that made the evening news: oil at $103, diesel at $4.65, European gas prices doubling. The brief closed with a structural claim: when the operating layer breaks at a chokepoint of this scale, the full consequences do not arrive at once. They propagate through supply chains at different speeds, and each week reveals dependencies that were invisible before the system broke.

The helium crisis is that claim, confirmed.

The same war risk insurance withdrawal that stopped oil tankers also stopped helium. Approximately 200 specialized cryogenic containers used to transport liquid helium were stranded in the Strait of Hormuz at the outset of the war, according to Phil Kornbluth, a former gas industry executive and helium industry consultant. Qatar, which produces roughly one-third of the world’s helium supply, has halted production after Iranian strikes damaged helium lines at Ras Laffan, the world’s largest LNG complex. QatarEnergy’s CEO and Rystad Energy estimate it could take three to five years to rebuild those lines. On March 17, Airgas, one of the largest packaged gas distributors in the United States and a subsidiary of Air Liquide SA, declared force majeure on helium shipments.

Helium is irreplaceable in the manufacture of advanced semiconductors. Without it, the $650 billion that Alphabet, Amazon, Meta, and Microsoft have committed to AI infrastructure in 2026 cannot be spent as planned. The AI chip supply chain, already strained by demand that exceeds manufacturing capacity, now faces a constraint that money alone cannot solve.

This is an intelligence update. We are in the fog of war. The full implications of the Hormuz closure will continue to emerge for weeks and months. The helium disruption is the latest, and it extends the operating layer failure from the gas pump to the data center.

Read more

The Strait of Hormuz, the California wildfire market, and the Florida property insurance crisis appear to be three different stories. One is a war. One is a climate catastrophe. One is a litigation epidemic.

They are the same story.

In each case, the private insurance market’s ability to assess, price, and transfer risk broke down. In each case, economic activity contracted or stopped. In each case, the government had to step in and become the insurer. The only variable is the speed at which it happened.

The Strait of Hormuz took 72 hours. California has been breaking for years. Australia is replicating the pattern across an entire continent. Florida took a decade to break and is now showing what recovery looks like.

The common failure in all four is what this publication calls the operating layer: the data infrastructure, risk models, institutional knowledge, and organizational capacity that sit between raw risk and an insurance price. When the operating layer works, insurance functions invisibly and economic activity flows. When it breaks, insurance withdraws and economic activity contracts to whatever can be self-funded.

This is not metaphor. It is the mechanism by which insurance operates as infrastructure, as essential to economic activity as roads, ports, and power grids, and as consequential when it fails.

Read more

The technology works. The data exists. The carriers bought it, deployed it, and announced it on stage at conferences. So why is more than half the industry not using what they paid for?

In this first episode, I lay out the framework for everything this show will cover. I call it the operating layer: the space between having data and getting value from it. The messy middle where models, technology, and capital meet actual organizations run by actual humans.

I walk through the Spectrum (theorists on one end, evangelists on the other, and why both are right but both fail), the Telematics Adoption Paradox, why the people who resist your technology are usually being perfectly rational, and what the companies that survived the operating layer did differently.

The patterns are drawn from insurance and telematics because that is where I have spent 20+ years. But the thesis is universal. If you are sitting on sensor data in any industry and you cannot figure out why the gold mine has not produced gold, this episode is about your problem too.

This episode includes a companion diagnostic scorecard: 10 indicators, five on the theorist side and five on the evangelist side. Score yourself 0 to 2 on each. Your totals tell you whether you are in the Theorist Trap, the Evangelist Trap, or the Messy Middle. The scorecard is a free PDF download inside the companion analysis below.

Companion analysis available at:

Market Intelligence reports available to subscribers at structuralsignal.com

©️2026 Indenseo Corporation

Published by Indenseo | structuralsignal.com

Executive Summary

The Internet of Things ecosystem deploys 21.1 billion connected devices inside a market valued at approximately $1.35 trillion, growing at 15% compound annual growth rate (Mordor Intelligence, November 2025). That is the infrastructure story. It matters because it tells the reader how much raw material exists and how fast it is growing. But the infrastructure number is not the market this report analyzes.

Read more

In Episode 1, I introduced the spectrum: on one end, the theorist who sees the inefficiency. On the other, the evangelist who sees the potential. Both are correct. Most of what they build fails in the middle. This companion piece turns that framework into a diagnostic you can apply to your own company, initiative, or investment thesis.

How to Use This

The spectrum is not a personality test. It is a structural assessment. Every data monetization initiative, every IoT integration project, and every technology adoption effort inside a legacy organization sits somewhere on it. The position determines the nature of the operating layer risk, and the nature of the risk determines what you need to fix.

Most failures are not caused by being too far toward one end. They are caused by misdiagnosing which end they are closer to, and therefore building the wrong capabilities to survive the middle.

Score each indicator below on a 0-1-2 scale:

0 = Does not describe your situation. 1 = Partially describes your situation. 2 = Strongly describes your situation.

Be honest. The value of this diagnostic is proportional to the honesty of the inputs.

Read more

Structural Signal is about that middle. Host Kevin Henderson calls it the operating layer: what actually happens when models, technology, and capital meet real organizations run by real humans.

Kevin draws on 20+ years of operational experience, from processing billions of miles of telematics data across multiple continents to building data monetization programs from zero to revenue.

If you are an insurance executive wondering why your telematics program still isn't delivering results, start here. If you are sitting on sensor data and can't figure out why the gold mine hasn't produced gold, this show is about your problem. If you want to understand the hidden markets being built on telematics and sensor data, we map out how they actually work every week.

Published weekly. Produced by Indenseo.

Between October and January, Anthropic signed enterprise partnerships covering more than one million employees across Deloitte, Cognizant, Accenture, and Allianz. In the same period, Accenture announced near-identical AI partnerships with both Anthropic and OpenAI, eight days apart, equipping tens of thousands of consultants with competing AI backends and parallel implementation playbooks. AIG put Anthropic’s CEO and Palantir’s CEO on stage together at its Investor Day, then launched a Lloyd’s syndicate with Palantir in December. State Farm signed on as a launch partner for OpenAI. Nationwide committed $1.5 billion in technology investment through 2028, including $100 million annually for AI.

The press releases all say “transform.” Read the fine print, and every one of these deals describes the same thing: giving existing employees access to AI tools that process existing workflows through existing business models, faster. This is not transformation. This is optimization. And the distinction matters – because we have seen this movie before.

The Ecosystem

A supply chain is forming around legacy carrier AI adoption, and each layer tells part of the story.

At the foundation sit the LLM providers, Anthropic and OpenAI, building the models and signing carrier deals. Anthropic now commands 40% of the enterprise LLM market, up from 12% in 2023, according to a December 2025 Menlo Ventures survey of approximately 500 U.S. enterprise decision-makers. (Menlo Ventures is an Anthropic investor, relevant context for evaluating the data.) Above the models sit analytics platforms like Palantir, whose Foundry platform now serves as AIG’s data backbone, accessing more than four million industry data points through an AIG-built ontology. Above that sits the consulting intermediary layer, Accenture, Deloitte, KPMG, PwC, mediating the implementation at senior consultant rates that industry benchmarks place at $250-$500+ per hour for AI-specific work. Then come the platform modernization vendors like Equisoft, embedding Claude directly into life insurance systems. And at the top sit the carriers themselves, Allianz, AIG, State Farm, Nationwide, Zurich, buying all of it.

Each layer extracts rent. None of them is transforming the underlying business model. Together they represent an enormous infrastructure of technology companies selling to legacy carriers, not competing with them.

The Picks and Shovels

That last point deserves emphasis, because it reveals something important about where insurance stands in the technology cycle.

Amazon did not partner with Borders. Google did not sell search tools to newspaper classified departments. Netflix did not consult for Blockbuster. They competed directly and won. The technology companies driving disruption in retail, media, and entertainment built alternatives that replaced incumbents. In insurance, the pattern reversed for the giants. Not long ago, the industry fear was that Google, Amazon, or another technology giant would enter insurance directly and disrupt carriers the way they disrupted retail and media. That did not happen. Some of the largest technology companies in the world looked at insurance and chose to be vendors. Anthropic, OpenAI, Palantir, and Accenture are all content to sell tools to legacy carriers rather than build competing insurance products. They view insurance as a customer base, not a market to transform.

The venture capital flowing into insurtech tells the same story. As I documented in Dot-Com Survivors and the Insurtech Parallel, most of the investment capital in insurance technology is going to B2B infrastructure: companies that will sell tools and platforms to carriers, not companies building full-stack alternatives that compete with them. Most of the capital structure of insurance technology, corporate and venture, has chosen the picks-and-shovels position.

The picks-and-shovels position is familiar from the dot-com era. Sun Microsystems sold servers to internet startups and to the legacy companies those startups were about to disrupt. Sun’s slogan was “We put the dot in dot-com,” but they also put servers in newspaper data centers. The hardware did not care who won. The analogy here is not to Anthropic’s technology, Claude is a fundamentally different kind of product than a SPARC server, but to the role the vendor ecosystem is playing. Infrastructure providers sell to everyone digging. They do not pick sides in the transformation battle.

Accenture’s dual partnerships make this explicit: OpenAI on December 1, Anthropic on December 9, selling the same consulting methodology with two different AI backends to whoever will buy. When the consulting layer is model-agnostic, the product is not the model. The product is the consulting fee. And the carriers buying it, whether they are building internally, hiring Accenture to build, or purchasing off the shelf, are still building on top of legacy systems. Allianz’s thousands of developers now have access to Claude Code. They are using it to write better code on the same architecture. The tool is powerful. The foundation it is being applied to is not new.

Optimized Satisficing

Satisficing, Herbert Simon’s term for choosing “good enough” over optimal, is the structural condition of legacy insurance. As I explored in Why Good Enough Is Killing Insurance for Carrier Management, carriers optimize for acceptable profits plus internal stability rather than pursuing the harder path of genuine transformation. AI does not change this dynamic. AI accelerates it.

Here is the structural reason. Large language models are trained on existing data. They are extraordinary at pattern recognition, finding regularities in the corpus they were trained on and applying those patterns at scale. Researchers describe them as “statistical models of language” that “rely on memorized patterns over genuine abstraction.” A landmark study on LLM idea generation found that “no evaluations have shown that LLM systems can take the very first step of producing novel, expert-level ideas.” The novelty LLMs produce is bounded recombination of known elements; impressive, but not the same as generating approaches that do not exist in the training data.

Deploy Claude at a legacy carrier and it does exactly what it was designed to do: optimize existing patterns. It processes the same submissions through the same underwriting model, faster. It extracts data from the same document formats with fewer errors. It matches risks against the same historical loss experience more efficiently. All real improvements. None of them transformative. AIG CEO Peter Zaffino described Anthropic’s impact in a TIME interview as allowing underwriters to get “quality insights from data in a fraction of the time” – and that is precisely the point. The insights come faster. The paradigm does not change.

A taxonomy of real-world LLM business model transformations found that actual deployments fall into “efficiency gains, service enhancement, and product extension,” not creation of fundamentally new business models. This is optimized satisficing: using the most powerful pattern recognition technology ever created to get better at the thing you already do, when the question is whether the thing you already do is the right thing to be doing at all.

The AIG Case Study

AIG’s technology partnership history, documented entirely from public record, illustrates this pattern across three consecutive eras.

September 2016: AIG, Hamilton Insurance Group, and Two Sigma launched Attune, a “data-enabled platform” targeting the $80 billion U.S. SME commercial insurance market. AIG CEO Peter Hancock called it “an important way forward for the insurance industry as it adapts to the disruptive forces of data analytics.” Hamilton CEO Brian Duperreault said Attune had “the potential to transform underwriting.” In February 2017, they hired OnDeck’s COO as Attune CEO. Duperreault called it “a perfect fit.” By May 2017, Duperreault, now AIG’s CEO, having moved from Hamilton, expanded the partnership and declared that “cross-industry partnerships, what’s now called insurtech, are the way to go” as the companies worked “to transform our industry.”

October 2021: AIG, Hamilton, and Two Sigma sold Attune to Coalition, a cyber insurance startup. All-cash transaction, terms undisclosed. Five years of press releases about “transforming the industry,” and they divested the entire thing.

April 2025: AIG’s Investor Day featured Anthropic CEO Dario Amodei and Palantir CEO Alex Karp alongside Zaffino. Zaffino, as quoted on Anthropic’s website, now declared their Anthropic partnership “will fundamentally transform how we approach underwriting at scale.” In December 2025, Karp praised AIG’s “forward-thinking approach to revolutionizing the insurance sector.”

The language was recycled almost verbatim. “Transform underwriting” in 2016 and 2017. “Fundamentally transform how we approach underwriting” in 2025. “Revolutionizing the insurance sector” from a new partner in the same breath. Meanwhile, AIG lost more than $30 billion in underwriting from 2008 to 2018 and invested more than $1 billion in data technology over five years.

The logic is simple. If the Two Sigma partnership had delivered transformative results, there would be no Palantir partnership. If the Palantir partnership were sufficient, Anthropic would not need to be on stage at Investor Day. The technology changes. The organizational constraints do not. That is satisficing: cycling through technology partnerships because each one is organizationally acceptable, even though none of them addresses the structural question.

The Cost Structure Tells the Story

The enterprise AI ROI data is convergent and more nuanced than the headlines suggest. MIT’s Project NANDA, a July 2025 study reviewing more than 300 AI initiatives across 52 organizations, found that 95% of enterprises are seeing zero measurable bottom-line impact from AI investment. McKinsey’s 2025 Global Survey found more than 80% of organizations report no tangible EBIT impact. These are midterm grades, not final verdicts. The adoption curve is early, and the results will evolve. But the pattern in the data already points somewhere specific.

BCG’s October 2025 research is the finding that matters most. Sixty percent of firms are generating “hardly any material value” from AI. But 5%, BCG calls them “future-built,” are producing 5x the revenue increases and 3x the cost reductions of everyone else. AI is not failing. AI is failing at organizations that deploy it as optimization of existing processes. The 5% that built their architecture around what AI makes possible are pulling away. The 95% bolting AI onto legacy workflows are not going to close that gap by doing more of the same thing: more consultants, more middleware, another partnership announcement. The deployment model is the variable, not the technology.

The MIT finding reinforces this. The 95% figure represents zero measurable P&L impact, not that the tools do not work. The report explicitly notes that AI tools “primarily enhance individual productivity.” The failure is in converting individual productivity gains into enterprise-level financial impact. This is the optimized satisficing thesis stated in research terms: the technology works exactly as designed. The organization cannot translate operational improvement into strategic transformation. And more spending will not solve an architecture problem.

For a carrier like Allianz or AIG, “deploying AI” means enterprise LLM licensing at 156,000-employee scale, consulting intermediaries billing at $300–$500+ per hour over multi-year engagements, analytics platforms like Palantir, cloud infrastructure, internal training and change management, governance and compliance systems logging every interaction, and ongoing legacy system integration through middleware. Seventy-six percent of AI use cases are now purchased rather than built internally, up from 53% in 2024, and 71% of U.S. insurers cite difficulties integrating AI with legacy systems. The consulting-mediated, bolt-on deployment model is not incidental. It is the primary delivery mechanism and the ROI data measures its results.

What This Actually Means

This analysis is not a critique of AI technology. Claude, GPT, and the models behind them are genuinely powerful. It is not a critique of Anthropic, Palantir, or Accenture. They are selling valuable products to willing buyers, exactly as Sun sold servers to newspapers and startups alike. And it is not an argument that these carriers are foolish. Regulatory barriers, capital requirements, and the fundamental complexity of underwriting risk have prevented pure technology companies from disrupting insurance from the outside, which is precisely why many companies in the technology ecosystem are selling to carriers rather than competing with them.

The argument is narrower and more specific. Deploying AI through the legacy carrier ecosystem, however impressive the technology, however large the investment, produces optimization of existing patterns, not transformation of the business model. The press releases say “transform.” The fine print says “faster processing of the same submissions through the same model.” The gap between the rhetoric and the reality is the story. And the early ROI data, BCG’s 5% pulling away from the other 95%, suggests the gap will widen, not close. The differentiator is not who spent the most on AI. It is who built the architecture that lets AI do what it is actually capable of.

The carriers deploying Claude to 156,000 employees, hiring Accenture to manage it, running it through Palantir, and training staff through vendor-led workshops believe they are transforming. They are optimizing. And as anyone who watched newspapers build websites in 1998 can tell you, confusing the two has consequences that take years to become visible – and by then, the organizations that built differently have already won.

This article was originally published on the Indenseo blog at indenseo.com/blog.

Author Note: This analysis draws on publicly available academic research, industry data, and regulatory filings. Statistics are cited to primary sources where available.

AI Disclosure: Research compilation utilized AI tools to discover and verify publicly available data sources and citations. All analysis, interpretation, and conclusions are original work.

Between 2020 and 2025, private equity firms deployed an estimated $40-50 billion into insurance distribution; more than any other insurance segment. As I examined in my recent analysis of PE’s insurance investment playbook, distribution attracts PE because the value creation is operational efficiency rather than operational transformation.

But the scale of what’s happening deserves closer examination. This isn’t just consolidation – it’s a fundamental restructuring of market power in insurance distribution.

The Concentration Story

The market’s consolidation trajectory tells a story the aggregate statistics miss.

In September 2021, 152 unique buyers competed for insurance agency acquisitions. By September 2025, that number had collapsed to just 96, a 37% reduction in market participants even as deal activity remained elevated.

The concentration at the top is even more striking. The top 10% of buyers controlled 56% of all deals by 2025, up from 46% four years earlier. Just three firms – BroadStreet Partners, Hub International, and Inszone Insurance Services – accounted for 25% of all transactions over trailing four quarters in early 2025.

BroadStreet alone completed 69 deals in 2025, 90 in 2024, and 59 in 2023, maintaining its position as one of the most active buyers for three consecutive years. This isn’t diversified competition; it’s platform accumulation by a shrinking group of increasingly dominant acquirers.

The Price of Scarcity

As buyer concentration intensified, valuations followed. Average EV/EBITDA multiples increased 27% between the 2019-2021 and 2022-2025 periods: a premium reflecting intensifying competition among fewer, larger buyers for quality assets.

The mega-deals have grown correspondingly. Arthur J. Gallagher announced the $13.45 billion AssuredPartners acquisition in late 2024: GTCR’s exit from a platform they had originally acquired in 2011, sold to Apax in 2015, reacquired in partnership with Apax in 2019, and built through hundreds of bolt-on acquisitions before this final exit.

These transactions aren’t transformational investments. They’re the logical endpoint of rollup economics: consolidate, optimize, exit to a strategic buyer. AssuredPartners was built through this exact playbook across multiple PE ownership cycles, and its sale to Gallagher, one of the world’s largest insurance brokerages, represents the terminal transaction PE sponsors are ultimately building toward.

The Platform Lifecycle

The AssuredPartners trajectory illustrates the distribution rollup lifecycle: GTCR acquired the platform in 2011, built it through acquisitions, sold to Apax in 2015, partnered with Apax to reacquire in 2019, continued building, then exited to Gallagher in 2024-2025. That final sale to a strategic buyer ends the PE cycle – Gallagher integrates AssuredPartners into permanent operations rather than positioning for another flip.

Each stage of this lifecycle creates returns for the PE sponsors involved. Each stage also increases the platform’s size and the multiple required for the next buyer to achieve similar returns. The question is where this progression leads when strategic exit is the only remaining option.

Over 35,000 independent agencies still operate in the US market. From PE’s perspective, this represents a nearly inexhaustible supply of acquisition targets. But as buyer concentration intensifies and multiples expand, the economics become increasingly challenging. The platforms that entered at 13x multiples face different math than those acquiring at 17x.

The Fragmentation Paradox

Here’s what makes the distribution pattern worth examining: Despite a decade of aggressive consolidation and $40-50 billion in capital deployment, the market’s fundamental structure persists. Those 35,000 independent agencies still exist. PE hasn’t consolidated the industry, it has created larger platforms that continue to feed on smaller targets.

This is rollup economics working exactly as designed. Complete consolidation isn’t the goal. The goal is building platforms large enough to command premium exit multiples from strategic acquirers or larger PE funds.

But the buyer concentration data suggests the end game may be approaching. With unique buyers declining from 152 to 96 and the top 10% controlling 56% of deals, the number of potential exit partners is shrinking even as platform sizes grow. At some point, the platforms become too large for PE-to-PE transactions and require strategic buyers: Gallagher, Aon, Marsh, Brown & Brown, willing to pay premium multiples for permanent integration.

The Gallagher-AssuredPartners deal is exactly this dynamic playing out. After three PE ownership cycles, the platform reached a scale where only a strategic buyer made sense. That’s not a failure of the model, it’s the model working as designed. But it does mean the rollup runway has an endpoint.

The question for investors evaluating insurance opportunities isn’t whether distribution rollups work. Clearly they do, and have generated substantial returns for the sponsors involved. The question is whether the current multiple environment and buyer concentration dynamics have already extracted most of the accessible value, or whether the fragmented agency base provides runway for continued platform building.

Understanding that dynamic helps clarify where different types of capital might find advantages: in distribution, or in segments PE systematically avoids.

This article was originally published on the Indenseo blog at indenseo.com/blog.

Author Note: This analysis draws on publicly available academic research, industry data, and regulatory filings. Statistics are cited to primary sources where available.

AI Disclosure: Research compilation utilized AI tools to discover and verify publicly available data sources and citations. All analysis, interpretation, and conclusions are original work.

This analysis is part of an ongoing series examining private equity’s approach to insurance investment and what revealed preferences tell us about market structure and opportunity.

Private equity has deployed over $150 billion into insurance since 2020. The thesis is straightforward: superior returns justify the illiquidity premium. But what does the performance data actually show?

The Kauffman Foundation’s analysis of its own venture and PE portfolio, one of the most comprehensive institutional self-assessments ever published, found that 62% of funds failed to exceed returns available from public markets after fees were paid. This wasn’t a sample of underperforming funds. It was the Foundation’s entire portfolio, including investments with “notable and exclusive partnership brands.”

The data indicates a persistent gap between gross and net performance: 78% of funds failed to deliver returns sufficient to reward patient, expensive, long-term investing. Only 20 of 100 funds beat a public market equivalent by more than 3% annually, and half of those began investing prior to 1995.

A January 2024 Harvard Business School working paper examining PE performance in the post-Global Financial Crisis era reinforced these findings, concluding that “the average or median PE funds do not actually outperform their PMEs since the GFC.”

The Fee Drag Problem

The standard 2-and-20 model, 2% annual management fees plus 20% carried interest on profits, compounds to significant value extraction over a fund’s life. Academic research from the Annual Review of Financial Economics estimates that fees consume 5-8% of annual gross returns.

The structural fee burden creates a high hurdle rate. A fund generating 15% gross returns might deliver only 8-10% net to LPs after fees, barely competitive with public market alternatives and substantially below what LP expectations require to justify the illiquidity premium.

For insurance investments specifically, this fee drag becomes particularly problematic. Genuine transformation in regulated industries requires 7-15 years. A 10-year transformation generating 15% gross returns, eroded by fees over that extended timeline, may deliver returns that don’t compensate for the capital lock-up.

Dispersion Matters More Than Averages

The data don’t suggest PE creates no value. Top-quartile funds do outperform, and the dispersion between top and bottom quartile is substantial, 14-18% annually according to Verus Investments research.

This dispersion is the critical insight. PE as an asset class doesn’t systematically outperform public markets. Individual manager skill and strategy fit determine returns. The wide performance spread means manager selection matters more than asset class allocation.

For LPs without access to consistently top-performing managers, the case for PE allocation weakens considerably relative to indexed alternatives. The question isn’t whether PE can generate returns; it clearly can in the right hands. The question is whether any specific LP has access to the managers who consistently deliver those returns.

The Selectivity Imperative

The 62% statistic isn’t an indictment of private equity. It’s a call for selectivity.

PE remains a viable strategy for investors with access to top-performing managers and opportunities that genuinely fit the financial engineering playbook. The managers who consistently outperform share specific operational capabilities: dedicated portfolio operations teams that implement operational platforming across holdings, proprietary deal sourcing through industry-specific networks, and structured value creation playbooks covering everything from procurement optimization to add-on acquisition integration. In insurance specifically, top performers leverage syndicated reinsurance structures and regulatory arbitrage strategies that require deep domain expertise most generalist funds lack.

For everyone else, the data suggest that indexed alternatives may deliver comparable or superior risk-adjusted returns without the illiquidity, complexity, and fee burden.

The burden of proof has shifted. PE allocation now requires demonstrating access to managers who can overcome the structural headwinds; not assuming that private markets inherently outperform public ones.

This performance reality shapes how PE approaches insurance. In Private Equity’s Insurance Playbook, I examined how these structural constraints explain PE’s revealed preferences; why $150 billion flowed into life insurance and distribution while commercial auto carriers were systematically avoided. The performance data here explain the “why” behind that pattern: when fee drag compounds over long transformation timelines, even competent execution can produce mediocre returns.

Different capital structures fit different opportunities. The data make that case with uncomfortable clarity.

This article was originally published on the Indenseo blog at indenseo.com/blog.

Author Note: This analysis draws on publicly available academic research, industry data, and regulatory filings. Statistics are cited to primary sources where available.

AI Disclosure: Research compilation utilized AI tools to discover and verify publicly available data sources and citations. All analysis, interpretation, and conclusions are original work.

Follow the money, and the pattern becomes unmistakable.

Between 2020 and 2025, private equity firms deployed over $150 billion into the insurance industry. Apollo completed its $29 billion merger with Athene. KKR acquired Global Atlantic and grew assets from $72 billion to $158 billion within three years. Aon paid $13.4 billion for NFP. PE-backed buyers completed 70-73% of all insurance agency M&A transactions during this period, with purchase multiples expanding from 13.1x to 16.7x EV/EBITDA.

The money flowed decisively into life insurance, insurance distribution, and specialty lines. What did PE systematically avoid? Commercial auto insurance carriers. Despite representing a substantial segment of the global P&C market, $199.9 billion in premiums (US market – $72 billion) and growing, research reveals “extremely limited direct PE acquisition activity in commercial auto insurance carriers or fleet insurance specialists” during the entire period.

This revealed preference tells us something important about how PE approaches insurance. Not because PE lacks capital or appetite for insurance; clearly it has both. But because some markets can be optimized through financial engineering, while others require something PE isn’t structured to deliver: genuine operational transformation.

Understanding this distinction is essential for any investor evaluating insurance opportunities.

Private Equity Investment Pattern - Where the Money Went

Sources: PitchBook, NAIC Capital Markets Bureau, S&P Global Market Intelligence, Indenseo Research

The Private Equity Insurance Playbook

Private equity’s insurance thesis is straightforward and well-articulated. Insurance provides predictable, long-duration liabilities that create stable capital for credit investments. PE-owned life insurers invest more heavily in alternative assets, $42.9 billion in Schedule BA investments at year-end 2024, up 17% from the prior year. The spread between what insurers pay policyholders and what they earn on invested assets generates returns that justify the acquisition premium.

The thesis works in life insurance and annuities because the transformation required is primarily financial. PE firms excel at asset-liability optimization, regulatory arbitrage through reinsurance structures, and deploying insurance float into higher-yielding credit investments. These are financial engineering competencies that transfer directly from PE’s core skillset.

Insurance distribution attracts PE for similar reasons. Brokerages generate predictable recurring commission revenue, 60-80% from renewals, with EBITDA margins of 25-35%. The transformation required is operational efficiency: consolidating back offices, implementing common technology platforms, and achieving purchasing scale. These are the rollup economics PE has perfected across dozens of industries.

But examine the leadership PE installs in its insurance platforms, and a pattern emerges. The executives running PE-backed insurance companies overwhelmingly come from legacy carrier backgrounds. This makes intuitive sense – PE firms want “industry expertise” and “credibility with regulators and rating agencies.” But it also reveals something about PE’s approach to insurance investment.

PE firms hire legacy executives because they’re satisficing for familiarity rather than optimizing for transformation. They want someone who knows the regulatory environment, has relationships with reinsurers and rating agencies, and can execute the financial engineering playbook without requiring the PE partners to deeply understand insurance operations. This is rational given PE’s business model, but it has consequences.

The consequence is that PE imports the same mental models and operational assumptions that created the problems in the first place. When your leadership comes from organizations that have struggled with technology integration for decades, you get the same technology struggles. This isn’t criticism of the executives, they’re executing what they know. It’s an observation about what the hiring pattern reveals about PE’s theory of change.

Satisficing at Every Level

Herbert Simon received the Nobel Prize in Economics for describing how organizations actually make decisions. His central insight: they don’t optimize. Instead, they “satisfice”, a portmanteau of “satisfy” and “suffice” describing the tendency to accept options that meet a minimum threshold rather than searching for the best option.

Optimizing requires identifying all possible alternatives, determining all consequences, and selecting the single alternative that maximizes the objective function. Satisficing sets an aspiration level, searches sequentially, selects the first alternative that meets the threshold, and stops immediately.

As I explored in a recent Carrier Management article on human capital efficiency, the satisficing framework explains organizational behavior that otherwise seems irrational. Companies acknowledge problems, commission studies, launch initiatives and change almost nothing. They’re not failing to execute. They’re executing exactly what their incentive structures reward: acceptable outcomes rather than optimal ones.

The PE approach to insurance investment exhibits satisficing at every level of the stack.

At the thesis level, PE frames insurance opportunities as financial arbitrage rather than operational transformation. The stated goal is to generate “spread-related earnings” and “regulatory and investment alpha,” as articulated in Brookfield’s 2025 Investor Day presentation and Athene’s investor presentations. This is satisficing on the financial metrics that LPs measure: IRR, multiple on invested capital, distributions, rather than optimizing for sustainable competitive advantage.

At the talent level, PE hires legacy carrier executives who can execute the financial playbook without requiring PE partners to deeply engage with insurance operations. This satisfices for familiarity and relationship continuity rather than optimizing for transformation capability.

At the infrastructure level, PE portfolio companies make technology choices optimized for exit rather than operational performance. Standard technology stacks, Microsoft enterprise environments, established core systems, familiar vendor relationships, reduce due diligence complexity for the next buyer. This satisfices for transaction convenience rather than optimizing for competitive differentiation.

At the timeline level, PE fund structures create inherent misalignment with transformation horizons. A typical PE fund has a 10-year life with 5-year investment periods and expectations of 3-5 year hold periods per investment. Genuine operational transformation in regulated industries: building technology platforms, achieving regulatory approvals, changing customer behavior, typically requires 7-15 years. Fund timelines satisfice for LP liquidity expectations rather than optimizing for transformation outcomes.

Financial Engineering vs. Operational Transformation

Financial engineering works when the underlying business generates acceptable returns that can be amplified through capital structure optimization, operating leverage, and multiple arbitrage. The PE playbook, acquire at 8x, improve margins 200 basis points, exit at 10x, works when the business is already profitable and the transformation required is primarily about efficiency rather than reinvention.

Operational transformation is different. It requires changing how an organization actually works: the technology architecture, the talent composition, the operational workflows, the customer relationships. It often requires sustained investment through periods of negative returns before improvement materializes. And it requires leadership that understands the operational domain deeply enough to make non-obvious decisions.

The distinction explains where PE succeeds in insurance and where it systematically avoids participation.

Life insurance works for PE because the transformation is primarily financial. The underlying product, long-duration liabilities, is well-understood. The operational complexity is in asset-liability management and regulatory optimization. PE firms can add value through superior investment management and capital efficiency without fundamentally changing how insurance is produced and delivered.

Distribution works for PE because the transformation is primarily operational efficiency at scale. The underlying economics, commission revenue on policy placement, are proven. PE can add value through back-office consolidation, purchasing leverage, and technology standardization. The transformation required doesn’t challenge the fundamental business model.

Commercial auto insurance requires something different. Twelve out of thirteen years of underwriting losses. Combined ratios above 100% despite 55 consecutive quarters of rate increases. The problem isn’t capital, the industry has plenty. The problem isn’t pricing, carriers have tried sustained rate increases without achieving profitability. The problem isn’t awareness, every rating agency, consultant, and industry analyst has documented the challenges.

What if the problem isn’t capital? What if it’s approach?

Commercial Auto: A Market Requiring Transformation

Commercial auto insurance presents a case study in transformation requirements that financial engineering cannot address.

Consider the mathematics. Commercial auto has recorded combined ratios above 100% for 12 of the past 13 years, meaning insurers have paid out more in claims and expenses than they earned in premiums, year after year. Commercial auto liability alone posted a $6.4 billion underwriting loss in 2024, according to AM Best. This isn’t a temporary dislocation. It’s a persistent structural condition that has resisted every conventional remedy the industry has attempted.

Commercial Auto Insurance Combined Ratio (2012-2024)

2020 – pandemic year – only profitable year in 13

Source: AM Best, NAIC, Indenseo Research

The structural challenges are well-documented: social inflation driving nuclear verdicts (median $23.8 million in 2023), vehicle repair cost inflation from advanced technology components that cost thousands to replace, acute driver shortages affecting both loss frequency and fleet operations, and deteriorating road infrastructure that contributes to accident severity. These aren’t cyclical problems that resolve through pricing discipline. They’re structural conditions requiring operational innovation.

AM Best maintains a Negative outlook on commercial auto, noting that the segment “struggles with a combined ratio that has been consistently over 100% despite rate increases.” S&P Global projects the commercial auto combined ratio will slowly deteriorate from 104.4% in 2026 to 106.3% by 2029, no return to profitability within the medium-term horizon.

The technology disconnect is particularly instructive. While 77% of commercial fleets use telematics systems, only 40% of insurers use that data in underwriting decisions. And here’s the revealing statistic: 70% of fleet managers report they don’t share telematics data with insurers, and 79% said they’ve simply never been asked.

This isn’t a technology problem. The hardware works. The data exists. The fleets are willing partners. As I detailed in Why Insurance Telematics Integrations Fail in Carrier Management, the barriers are operational: legacy IT architectures that can’t ingest real-time data, workflow processes that don’t incorporate telematics insights, and organizational structures that fragment responsibility across underwriting, claims, and risk management silos.

Digital transformation projects in insurance carry failure rates exceeding 70%. The failure mode is consistent: technology implementations are approached as IT projects rather than operational transformations. Organizations buy new tools and try to bolt them onto unchanged workflows, the “Ferrari engine in a covered wagon” pattern that produces expensive disappointments rather than competitive advantage.

Solving commercial auto requires changing how risk is assessed, priced, and managed, not just adding technology to existing processes. It requires leadership with deep operational understanding of both insurance and the fleet ecosystem. It requires patient capital willing to support multi-year transformation without pressure for near-term exits.

PE’s revealed preference, systematically avoiding commercial auto carrier investments, suggests recognition that the playbook doesn’t fit. This isn’t a criticism of PE decision-making. It’s rational given PE’s business model and success metrics. But it reveals the boundaries of financial engineering as an approach to value creation.

The Structural Mismatch

The performance data raise questions about PE’s value proposition even in segments where the playbook theoretically applies.

The Kauffman Foundation’s analysis of its own venture and PE portfolio, one of the most comprehensive institutional self-assessments ever published, found that 62% of funds failed to exceed returns available from public markets after fees were paid. A January 2024 Harvard Business School working paper examining PE performance in the post-Global Financial Crisis era concluded that “the average or median PE funds do not actually outperform their PMEs since the GFC”.

These findings don’t suggest PE creates no value, top-quartile funds do outperform, and the dispersion between top and bottom quartile is substantial (14-18% annually). But they challenge the assumption that PE as an asset class systematically outperforms public alternatives after accounting for fees, illiquidity, and risk.

The fee structure matters for insurance investors specifically. The standard 2-and-20 model – 2% annual management fees plus 20% carried interest on profits -compounds to significant value extraction over fund life. Academic research estimates that fees consume 5-8% of annual gross returns, transforming potential outperformance into mediocrity or underperformance for many funds.

For insurance investments with long transformation horizons, this fee drag becomes particularly problematic. A 10-year transformation generating 15% gross returns might deliver only 8-10% net to LPs after fees; barely competitive with public market alternatives and substantially below what LP expectations require to justify the illiquidity premium.

The question isn’t whether PE firms are competent, they clearly are within their domain. The question is whether the PE model aligns with the requirements of specific insurance opportunities. When the playbook fits, financial engineering, operational efficiency, rollup economics, PE has demonstrated ability to create value. When the opportunity requires something different, patient capital, deep operational transformation, timeline horizons beyond fund life, the structural mismatch becomes limiting.

Implications for Different Investor Types

This analysis suggests different frameworks for different investor categories evaluating insurance opportunities.

For institutional LPs allocating to PE funds: Selectivity matters more than access. The wide dispersion between top-quartile and bottom-quartile performance (14-18% annually) means manager selection determines outcomes more than asset class allocation. The data suggest that PE as an asset class doesn’t systematically outperform public markets, individual manager skill and strategy fit determine returns. For LPs without access to consistently top-performing managers, the case for PE allocation weakens considerably relative to indexed alternatives.

For family offices considering direct insurance investment: The shift from intermediated to direct deployment may offer structural advantages in insurance. Family offices have increased direct investments to 17% of portfolios while fund investments decreased to 10%, a revealed preference suggesting those who can choose are increasingly choosing to bypass traditional fund structures. Co-investment programs at large institutional investors report net IRRs around 27% versus approximately 19% for traditional primary fund commitments, a differential driven largely by eliminating the fee structures that erode returns.

Patient capital with longer time horizons can support transformation timelines that fund structures cannot accommodate. The economics are particularly compelling for opportunities that require 7-15 year development periods, timelines that create structural misalignment with standard PE fund lives but align naturally with family office investment horizons.

For strategic investors and carriers: The PE playbook’s limitations in transformation-requiring segments may create opportunity for different approaches. Markets like commercial auto, systematically avoided by PE despite their scale, might reward operators with technology-first capabilities, patient capital backing, and willingness to pursue multi-year transformation without exit pressure. The infrastructure that PE finds unsuitable for their model may be precisely what creates sustainable competitive advantage for operators with different capital structures.

For entrepreneurs and operators in insurance: Capital structure matters as much as capital access. Raising PE fund capital for a transformation opportunity creates structural misalignment from day one, the fund’s timeline requirements will pressure decisions that may undermine long-term value creation. The insurance industry has seen this pattern repeatedly: PE-backed insurtechs achieving rapid growth metrics while destroying economic value, then failing when the underlying unit economics never materialized.

Matching capital structure to business characteristics isn’t about ideology, it’s about setting up conditions for success. Some opportunities genuinely fit the PE model and benefit from PE’s operational playbook and value-creation capabilities. Others require patient capital, and raising the wrong type of capital is itself a form of satisficing, accepting the first funding that meets the minimum threshold rather than optimizing for the capital structure that maximizes probability of success.

The framework isn’t “PE bad, alternatives good.” It’s more nuanced: different capital structures fit different opportunities. PE’s approach works where financial engineering and operational efficiency create value on fund-life timelines. It doesn’t fit where genuine transformation requires patient capital and deep operational commitment.

The Revealed Preference

Private equity’s insurance investment pattern over the past five years reveals both capabilities and limitations. The $150+ billion deployed into life insurance, distribution, and specialty lines demonstrates PE’s genuine appetite for the sector and ability to create value where the playbook fits.

The systematic avoidance of commercial auto carriers, despite the segment’s substantial size and obvious transformation potential, reveals the boundaries of that playbook. PE firms are sophisticated enough to recognize when opportunities don’t fit their model. They’re rational in declining to pursue investments requiring capabilities and timelines their fund structures can’t support.

For investors evaluating insurance opportunities, the analysis suggests moving beyond capital availability as the primary consideration. The insurance industry has access to substantial capital across PE funds, family offices, reinsurance relationships, and public markets. Capital isn’t the scarce resource.

What’s scarce is alignment between capital structure, transformation timeline, and operational capability. Markets requiring genuine transformation, not just financial optimization, need investors willing to match their approach to the actual requirements of value creation.

The question for any specific insurance opportunity isn’t whether capital is available. It’s whether the capital’s structure, timeline expectations, and operational engagement model match what the opportunity actually requires.

Different answers lead to different outcomes.

Author Note: This analysis draws on publicly available academic research, industry data, and regulatory filings. Statistics are cited to primary sources where available.

AI Disclosure: Research compilation utilized AI tools to discover and verify publicly available data sources and citations. All analysis, interpretation, and conclusions are original work.

This analysis is part of an ongoing series examining private equity’s approach to insurance investment and what revealed preferences tell us about market structure and opportunity

Every conversation about economic development focuses on the same institutions: capital markets, infrastructure investment, trade policy, central banks. Insurance rarely makes the list.

This is a mistake.

Commercial property and casualty insurance operates as foundational infrastructure, as essential to economic development as roads, ports, and electrical grids. When businesses can transfer catastrophic risk to insurers, they invest, hire, and grow. When they cannot, economic activity contracts to what can be self-funded and self-insured.The global “protection gap” – the difference between total economic losses and insured losses exceeds $1.8 trillion annually. This gap represents unrealized economic potential: businesses that don’t form, investments that don’t occur, jobs that don’t exist. For investors who ask what good their capital does in the world, commercial insurance offers an unusually clear answer—and an unusual alignment between impact and returns.

The Risk Transfer Enablement Thesis

The relationship between insurance and economic development operates through a simple mechanism: risk transfer enables risk-taking.

Consider a logistics company contemplating expansion. Without adequate insurance, the owner faces the prospect of total business loss from a single catastrophic accident. The rational response is conservative operation: limited growth, minimal debt, constrained geographic reach. With proper coverage, the same owner can invest in additional vehicles, hire drivers, and expand routes, confident that a serious accident will not destroy the enterprise.

This mechanism operates across commercial sectors. Contractors require liability coverage to bid on construction projects; larger projects require larger limits. Equipment breakdown coverage enables capital investment in expensive manufacturing machinery. Errors and omissions coverage enables professional services firms to serve larger clients with greater exposure. Crop insurance enables farmers to plant higher-value crops and invest in land improvement. Commercial auto coverage enables fleet formation and operation.

The academic evidence supports these observations. Three competing hypotheses emerged in the literature on the insurance-growth nexus. The “supply-leading hypothesis” argues that insurance development precedes and causes economic growth. The “demand-following hypothesis” suggests insurance development is a consequence rather than cause of growth. But the “feedback hypothesis” has emerged as dominant. Insurance and growth reinforce each other in a virtuous cycle. Recent econometric studies using sophisticated approaches reveal this bidirectional relationship is non-linear, with mutual reinforcement strongest at early stages of development.

The practical implication: developing economies need not wait for growth to drive insurance development. Deliberate expansion of insurance infrastructure can accelerate growth, particularly in commercial lines where business investment decisions are directly conditioned on coverage availability.

Commercial Auto: The Circulatory System of Modern Commerce

Within commercial insurance, commercial auto occupies a unique structural position. In the United States, trucking moves approximately 72% of the nation’s freight tonnage and generates over $906 billion in annual revenue. The industry employs 8.4 million people, including 3.5 million truck drivers; more than any other single occupation in many states.

Commercial vehicles are not merely tools; they are the circulatory system of modern economies. Without functional commercial fleets, supply chains collapse. The COVID-19 pandemic illustrated this starkly: commercial trucking was classified as “essential infrastructure” because modern commerce literally cannot function without it.

Yet commercial auto insurance has operated in crisis for over a decade. The sector has posted combined ratios above 100%, meaning insurers pay out more in claims and expenses than they collect in premiums, in 12 of the last 13 years [1]. Commercial auto insurance premiums have increased for 55 consecutive quarters [1]. That’s nearly 14 years of uninterrupted price hikes affecting every business that operates vehicles.The traditional insurance industry has essentially one tool: raise rates across entire categories. If you’re a plumber with a perfect safety record, your rates still increase because other plumbers are filing claims. You’re punished for risks you didn’t create.

The Small Fleet Penalty

The most striking structural inequity in commercial auto insurance is what industry analysts call “the small fleet penalty.”According to American Transportation Research Institute data, small carriers pay approximately 21.0 – 21.3 cents per mile in insurance costs, compared to 10.2 cents for large carriers, a more than 100% penalty. This disparity exists because they do not have access to tools to manage risks to reduce insurance costs that larger fleets do. Without granular data, insurers default to pricing small fleets at the worst end of their category’s experience.

Table 1: Small Fleet Insurance Cost Penalty

Source: American Transportation Research Institute, Operational Costs of Trucking (2025)

The structural consequences are severe. An 11 cent per mile cost disadvantage erodes the margins necessary for small business growth. In down markets, large carriers survive on thinner margins while small carriers with high fixed insurance costs are forced to exit. The rate increase cycle becomes self-reinforcing: as small operators exit, the remaining pool becomes less competitive, inviting further rate increases.

For entrepreneurs attempting to build businesses through fleet ownership, this represents a structural barrier to wealth creation. The pathway from owner-operator to fleet owner requires capital accumulation. When insurance costs consume disproportionate margin, that accumulation becomes nearly impossible.

Proxy-Based Underwriting and Economic Mobility

Beyond the small fleet penalty, commercial auto insurers rely heavily on “proxy-based underwriting”—using variables that correlate with risk but may also embed socioeconomic biases.

Credit-based insurance scores represent the most consequential proxy. A Consumer Federation of America study found that good drivers with poor credit pay premiums 115% higher than good drivers with excellent credit in personal auto; similar dynamics apply commercially. Credit scores correlate with past economic circumstances, not driving ability. An entrepreneur who experienced financial hardship, job loss, medical emergency, economic disruption, carries that history into their insurance pricing regardless of how safely they operate their fleet.

Geographic rating compounds the effect. Garaging location affects premiums, and businesses in urban areas face higher base rates regardless of individual driving patterns. Experience requirements penalize new authorities and young businesses, creating surcharges that disproportionately affect entrepreneurs entering the market.

The cumulative effect: the insurance market systematically disadvantages economic mobility. Operators with established businesses, accumulated capital, and stable credit histories receive preferential pricing. New entrants, small operators, and entrepreneurs rebuilding from adversity face compounded penalties.

For social impact investors, this market structure represents both a problem and an opportunity.

The Telematics Solution: Pricing Behavior Instead of Demographics

Telematics, technology that monitors driving behavior, vehicle usage, and safety metrics, offers a mechanism to fundamentally restructure commercial auto insurance pricing.

The evidence base is unusually strong. There is a 20-30% crash reductions when telematics is properly implemented. Multiple controlled fleet trials and longitudinal evaluations find that telematics‑based monitoring and coaching found a 40-60% reduction in risky behavior like speeding and harsh braking when embedded in a structured safety program. Longitudinal studies show behavioral changes persist beyond initial adoption periods. Large-scale observational data validates the link between telematics-measured behaviors and actual crash outcomes.

Usage-based insurance utilizing telematics data can decouple rates from demographic proxies. The mechanism assesses risk based on actual driving behavior, speeding, hard braking, cornering, hours of service compliance, rather than credit scores or zip codes. A driver with a low credit score but excellent safety habits can prove their low-risk profile directly to the insurer, bypassing proxy variables that embed historical disparities.

Market examples validate the approach. Companies like Cover Whale focus on small fleets (1-25 units), provide dashcams and AI-driven coaching, and offer coverage to new ventures. HDVI offers “dynamic pricing” where monthly premiums adjust based on safety scores; case studies indicate fleets can save up to 30%.

For a single-truck owner-operator, the financial impact compounds meaningfully. For a truck running 100,000 miles a year, that’s an additional $11,000 in annual insurance costs that could be retained earnings through telematics-enabled cost reduction can pay down debt faster, provide down payment for a second truck, or create reserves preventing business failure during market downturns.

The Developing Market Opportunity

The protection gap is not evenly distributed. It concentrates in emerging markets where insurance penetration remains low despite expanding economic activity.

When a natural disaster strikes an uninsured economy, reconstruction requires diversion of development capital. When an uninsured business suffers a catastrophic loss, employees lose jobs without unemployment coverage, suppliers lose customers without accounts receivable protection, and communities lose tax base.

The relationship between income and insurance penetration follows predictable patterns. Economies with GDP per capita below $5,000 typically show less than 1% non-life insurance penetration. The critical transition occurs in the $5,000-15,000 range, precisely where many African and Asian economies now sit. This is the window of maximum impact for insurance market development.

Commercial lines have particular importance in developing economies. International investors require adequate local insurance capacity as a condition of investment. A World Bank analysis finds that countries with higher insurance penetration tend to have smaller infrastructure investment gaps, suggesting that a developed insurance sector can play a meaningful role in mobilizing long‑term capital for infrastructure.. Export-oriented businesses require credit insurance, marine cargo coverage, and trade credit facilities. Small and medium enterprises drive employment growth in developing economies, and commercial coverage requirements often serve as the trigger for formal registration, banking relationships, and regulatory compliance.

Africa: The World’s Largest Untapped Commercial Insurance Opportunity

The African insurance market represents the world’s largest untapped opportunity for commercial insurance development. As of 2024, the total market value reaches approximately $92.9 billion, projected to grow to $160.9 billion by 2033.

The structural challenge is stark: South Africa dominates, representing nearly 70% of continental premiums with a penetration rate of 11.54%, comparable to OECD markets. The rest of the continent offers structural growth potential with penetration rates rarely exceeding 3%. Nigeria shows approximately 0.3% penetration despite being Africa’s most populous nation. Kenya operates at approximately 2-3%, depending on the year and source, serving as the continental innovation laboratory.

Table 2: African Insurance Market Penetration

Source: Swiss Re, KPMG Insurance in Africa Report

Technology is transforming insurance access in these markets. Mobile money platforms enable premium collection without banking infrastructure. Parametric products pay automatically when defined triggers occur without requiring loss adjustment infrastructure. Alternative data underwriting uses mobile phone usage patterns, utility payment history, and satellite imagery to provide risk assessment where traditional sources are unavailable.

Commercial auto in Africa presents unique dynamics. South Africa leads telematics adoption with 2.3 million units installed in 2023, projected to reach 3.8 million by 2028. Companies like Karooooo (formerly Cartrack) and Netstar are expanding northward. Starlink and other satellite providers are solving connectivity infrastructure challenges that previously limited telematics deployment outside major urban centers.

For impact investors seeking structural growth opportunities, African commercial insurance development represents a multi-decade trajectory with compounding returns and measurable development impact.

Why Impact and Returns Are Aligned

The critical insight for social impact investors: in commercial auto insurance, impact and returns move in the same direction.

The mechanism is straightforward. Telematics identifies safer drivers, which generates lower claims costs, which produces better margins. Lower premiums for good risks create competitive advantage, which drives market share growth. Safer fleets mean fewer accidents, which saves lives.

Insurance companies with superior risk selection consistently outperform those that simply price for expected catastrophe. Impact doesn’t sacrifice returns, it generates them.

This is not a situation where impact is a separate department running philanthropic programs alongside the profit center. The same data that improves safety outcomes improves underwriting margins. The metrics that demonstrate impact, accident reduction, behavior change, are the same metrics that demonstrate underwriting performance. If accidents don’t decrease, claims costs don’t decrease. If claims costs don’t decrease, margins suffer.

Unlike many impact investments where outcomes are hard to measure or take decades to materialize, telematics-enabled insurance produces measurable results within months: fewer harsh braking events, reduced speeding, lower accident rates. The feedback loop is tight enough to demonstrate causation, not just correlation.

The Scale of Potential Impact

The potential scale merits consideration:

If telematics achieves a 25% reduction in the 5,000+ annual U.S. commercial vehicle deaths, that’s 1,250+ lives saved per year. Apply similar reductions to the 70,000+ injuries, and 17,500+ people are spared serious harm annually.

Closing the more than 100% cost gap for small fleets represents billions in annual savings industry-wide. That capital, retained by small business owners instead of extracted by insurers, compounds into business equity.

The $1.8 trillion global protection gap represents an equivalent scale of unrealized economic potential: businesses that could form, investments that could occur, jobs that could exist if insurance infrastructure developed appropriately.

A Narrative Worth Telling

For investors evaluating impact investments, the insurance sector offers a story that resonates:

“We invested in a company that makes roads safer by using data to identify and reward good drivers. The traditional insurance industry just raises everyone’s rates when accidents increase. This approach actually prevents accidents and the data proves it works. Along the way, it gives small business owners a fair shot at affordable insurance based on how they actually drive, not their credit score or zip code. The business model aligns incentives: safer fleets mean fewer claims, which means better margins. Impact and returns move in the same direction.”

The combination of immediate measurability, clear attribution, and incentive alignment makes commercial auto insurance unusually well-suited for impact investors who want accountability, not just aspirations.

What This Research Implies

The evidence synthesized here points to specific characteristics that would define an effective telematics-enabled commercial auto insurance model:

Risk Selection: The ability to identify safer-than-average risks within categories that traditional insurance prices as homogeneous pools. The more than 100% cost penalty for small fleets exists because the market lacks risk management and cost reduction tools for small fleets.

Loss Prevention: Active engagement in reducing accidents, not just pricing for expected losses. The research is clear that passive telematics, hardware without feedback, produces negligible safety improvement. Effective models require closed-loop systems with real-time feedback and coaching.

Accessible Pricing: Mechanisms to serve markets that traditional insurance has effectively abandoned through pricing. Small fleets, new entrants, and entrepreneurs rebuilding from adversity face structural barriers that telematics can address by replacing proxy-based underwriting with behavior-based assessment.

Measurable Outcomes: Commitment to tracking and reporting impact metrics, not as a marketing exercise, but as core operational data.

The mechanisms described are not U.S.-specific. Telematics technology is globally deployable. The small fleet penalty exists wherever insurance markets rely on class-based rather than individual pricing. Developing markets with low insurance penetration represent structural growth opportunities.

Commercial property and casualty insurance rarely appears in discussions of economic development infrastructure. The evidence suggests it should.

Author Note: This analysis draws on publicly available academic research, industry data, and regulatory filings. Statistics are cited to primary sources where available.

AI Disclosure: Research compilation utilized AI tools to discover and verify publicly available data sources and citations. All analysis, interpretation, and conclusions are original work.

Endnotes

[1] Indenseo Research analysis.

The devastation was real. Between March 2000 and October 2002, the NASDAQ collapsed 78%, erasing approximately $5 trillion in market value. The technology sector shed 540,000 jobs in just two years. In Silicon Valley, unemployment surged from 1.3% to 8%, and nearly one in four unemployed workers remained jobless for more than 27 weeks. Santa Clara County lost more than 200,000 jobs – the largest decline for any metropolitan area since the Great Depression.

I arrived at Infoseek in the mid-1990s during the bubble, then left for a startup that would later appear on F***edCompany.com’s failure tracker. Infoseek would end up on that list too, after Disney shut it down. I watched colleagues leave one dying company only for their next company to die, sometimes repeating this pattern three or four times. Some eventually left tech entirely – they couldn’t take another company dying under them.

At an east coast dinner party that I attended during the worst of it, a Fortune 500 CFO declared that dot-com companies “were not real companies” and the people who worked at them “were not real business people.” “Real companies,” he insisted, would never hire them. That CFO’s company was subsequently acquired; its brand name has since disappeared. Amazon’s market cap now exceeds $2 trillion.

Legacy carriers celebrating insurtech failures today are making the same analytical error. The pattern is identical: massive crash, schadenfreude from incumbents, quiet building by survivors, eventual market transformation. The only question is whether you recognize the pattern early enough to act on it.

No Bailouts, Real Consequences

Unlike 2008’s bank bailouts or 2020’s pandemic relief, the dot-com crash received no government intervention. Nobody felt sorry for tech workers who had gambled on stock options. Pets.com’s $300 million, Webvan’s $800 million, Excite@Home’s $35 billion – all vaporized with no safety net.

The cruelty extended to employees who thought they’d done everything right. The Alternative Minimum Tax trapped workers who exercised stock options at peak valuations, then watched prices collapse. They owed six-figure or seven-figure tax bills on gains they never realized, on stock now worth pennies. Some filed personal bankruptcy specifically because of stock option tax obligations. One local executive purchased 240,000 shares at $0.14 when the stock traded at $19.75 – $4.7 million in paper value for $33,700. By year-end the shares were worth $45,600, but he faced a tax bill exceeding $1 million.

Drive through Silicon Valley today and the landscape tells the destruction story. A former Sun Microsystems satellite office, where Oracle briefly hung a sign over the old logo before shutting it down, is now Facebook’s headquarters. The former Silicon Graphics headquarters, which I visited during my Infoseek days when SGI dominated CGI for film and television, is now Google’s headquarters. Both companies failed for the same reason: no answer to the x86 chip. The buildings survived. The companies that couldn’t adapt didn’t.

But this matters: because there were no bailouts, creative destruction actually worked. Bad companies and bad management teams faced consequences. Investors who bet on unsustainable models lost their money. Capital was reallocated from failures to survivors. The painful clearing created the foundation for what came next.

The Dot-Com Timeline: From Mania to Dominance

The NASDAQ peaked at 5,048.62 on March 10, 2000. It bottomed at 1,114.11 on October 9, 2002 – 30 months of destruction. Amazon’s stock crashed 94%, from $107 to $6. Priceline fell 99%, from $974 to $6.60. Cisco lost 90% of its value and didn’t recover its March 2000 peak until December 2025, 25 years later.

Then came the quiet period. From 2002 to 2005, while incumbents relaxed and the media wrote post-mortems on the “dot-com folly,” the survivors built. Amazon posted its first profitable quarter in Q4 2001 – during the depths of the crash. Rather than cutting, Bezos expanded into new categories and began developing the infrastructure that would become AWS. Google went public in August 2004, four years after the peak and two years after the trough, proving that world-class businesses could launch in the aftermath of a crash.

The NASDAQ didn’t return to its March 2000 level until April 2015, 15 years later. But the survivors didn’t just recover; they obliterated their incumbents. Amazon’s market cap now exceeds Walmart’s by multiples. Google and Facebook captured the digital advertising revenue that newspapers thought they’d protected by surviving the crash. Netflix, founded in 1997 and nearly sold to Blockbuster for $50 million in 2000, introduced streaming in 2007 and now dominates global entertainment.

The Insurtech Timeline: Where We Are Now

Global insurtech funding rose from approximately $1.5 billion in 2015 to a peak of $15.8 billion in 2021 – a single year that accounted for more capital than the previous five years combined. The public market euphoria was even more extreme. Lemonade peaked at $188 per share in January 2021, valuing the company at over $10 billion despite less than $100 million in revenue. Root peaked at $486 (split-adjusted) in late 2020.

Overlay the timelines. Dot-com: bubble inflation 1995-2000 (5 years), peak March 2000, crash 2000-2002 (2.5 years), quiet period 2002-2007. Insurtech: bubble inflation 2015-2021 (6 years), peak 2021, crash 2022-2024 (2-3 years). We are now approximately 3-4 years post-peak – the equivalent of 2003-2004 in the dot-com cycle.

This is the quiet period. And the survivors are building.

Survivor Patterns Across Eras

The dot-com survivors shared specific characteristics. Amazon achieved profitability during the crash and used the quiet period to build AWS rather than retreat. Google focused on engineering headcount and infrastructure while competitors cut R&D. Netflix continued investing in its DVD unit economics while developing the streaming capability that would launch in 2007. All prioritized unit economics over growth-at-all-costs. All had founder leadership through the crisis. All treated the crash as an opportunity to build capabilities rather than a signal to harvest.

The insurtech survivors display identical patterns. Root Insurance achieved its first full year of GAAP profitability in 2024 – $31 million net income and $112 million adjusted EBITDA – with gross loss ratios improving from over 90% during its crisis years to approximately 59% in late 2025. The company didn’t just survive; it fundamentally restructured its underwriting and abandoned shotgun marketing for embedded distribution. Founder Alex Timm remains CEO, a rarity among 2015-era insurtechs.

Coalition’s gross written premiums exceeded $630 million in 2023, up roughly 15% year‑over‑year, and its July 2022 funding announcement cited over $775 million in run‑rate Gross Written Premium (GWP). maintaining a $5 billion valuation through its “Active Insurance” model that combines security software with risk transfer. Its loss ratios consistently outperform industry averages because it actively scans policyholders and prevents claims before they happen – a capability legacy carriers physically cannot replicate without rebuilding their tech stacks.

Next Insurance’s $2.6 billion acquisition by Munich Re validated the digital small business thesis. Shift Technology processes claims and documents at scale across 115+ insurance customers worldwide, with a SaaS delivery model distinct from traditional risk-bearing underwriting economics. Newfront’s acquisition by WTW for up to $1.3 billion proved that modernizing the brokerage model was faster than replacing it.

The Pattern Holds: B2B Infrastructure Outperforms Capital-Intensive Carriers

B2B insurance infrastructure software achieves 70-80% gross margins, substantially exceeding the operating economics of capital-intensive insurance carriers, which face combined ratios of 97-102% and net profit margins below 3% on underwriting alone (with some health insurers and new insurtechs showing negative net margins when investment income is excluded).

Commercial lines insurance commands significantly higher premiums than personal lines – middle market commercial policies typically range from $50,000 to $1,000,000 annually, compared to personal auto insurance averaging $2,300-$2,677 for full coverage and $627-$916 for minimum liability coverage. This premium variance creates superior economics for intermediaries focused on commercial segments, which often generate 10-15x greater commission income per policy than personal lines.

Embedded insurance distribution achieves three to four times lower customer acquisition costs compared to direct-to-consumer models, reducing traditional insurance CAC of $487–$1,280 per customer to an estimated $122-$300 range by leveraging contextual sales at the point of transaction. This distribution advantage, combined with infrastructure-like unit economics, explains why platforms prioritize embedded partnerships over direct distribution.

Strategic and patient capital from family offices (increasingly allocating to insurance for uncorrelated returns), reinsurers (backing a record 51 insurtech investments in Q3 2025), and incumbent carriers replace traditional Silicon Valley venture capital, which has retreated due to structural timeline misalignment – VC funds with 5-7 year lifecycles cannot accommodate insurance businesses requiring 7-10 years to achieve sustainable profitability. Public insurtech exits declined 79% since 2021, with only six exits recorded in 2024, forcing founders toward strategic and family office investors with permanent capital structures.

The Infrastructure Provider Warning: Sun Microsystems and Nvidia

I didn’t just observe Sun Microsystems from the outside. Infoseek was Sun’s second-largest customer after Intel. As head of content and technology licensing, I worked with Sun constantly. There were waiting lists for Sun’s latest and most powerful servers – actual waiting lists, not marketing gimmicks. Your position in Sun’s allocation queue signaled your importance in the technology hierarchy. Getting a top server meant you mattered.

Sun’s market cap peaked above $250 billion. The company’s slogan was “We put the dot in dot-com.” Its products were genuinely revolutionary – Java, Solaris, SPARC. Its customers were real. Its profits were real. At peak, Sun reported $2 billion in net income.

Years later at @Road – one of the last companies to IPO before the crash, and a survivor – the first data deliveries to telematics data customers were still powered by Sun servers. I watched Sun across the entire arc: essential infrastructure provider with waiting lists, the company everyone needed, sold to Oracle in 2010 for $7.4 billion. A 97% discount from peak. The company that powered the boom couldn’t survive the bust.

The demand collapse was the trigger, but not the full story. Sun had scaled operations to meet unsustainable demand from customers who subsequently vanished – when dot-coms failed en masse, they stopped buying servers. But the deeper failure was strategic: Sun had optimized for a high-margin proprietary hardware world that was disappearing. While Sun defended its SPARC architecture and premium pricing, cheaper x86 servers from Dell and HP captured the market from below. When Amazon launched EC2 in 2006, the “buy servers” model itself became optional. Sun’s “essential infrastructure” didn’t just lose customers – it became expendable. Being essential during a bubble provides no immunity when demand collapses and cheaper alternatives emerge simultaneously.

Anyone watching AI infrastructure today recognizes the echoes. Nvidia GPU allocation scarcity. Status signaling based on your position in the queue. Jensen Huang deciding who gets chips. The waiting list parallels are uncomfortable. This doesn’t predict Nvidia’s fate – the business model differences matter, the customer base is stronger, AI applications may prove more fundamental than dot-com applications. But the Sun lesson isn’t just about demand collapse. It’s about what happens when you’ve optimized for a world that might not exist in five years. The questions that matter: Are custom chips from hyperscalers – Google’s TPUs, Amazon’s Trainium, Microsoft’s Maia – the x86 of this cycle? Is inference-as-a-service the cloud computing equivalent? Nvidia may have excellent answers. But “essential” has never meant “safe.”

The False Survivor Warning

Surviving the crash is necessary but not sufficient. Yahoo survived the dot-com bust with massive market cap and user base intact. The company reported three consecutive profitable quarters by early 2003. Wall Street celebrated the “adult supervision” of CEO Terry Semel.

But Yahoo’s survival strategy relied on viewing the internet as a content distribution pipe rather than a technology platform. In the mid-2000s, Yahoo CEO Terry Semel publicly acknowledged that Yahoo had missed the opportunity to acquire Google when the company was smaller and less expensive By 2008, Microsoft offered $44.6 billion for Yahoo – and Yahoo rejected it, believing its “media platform” had immense latent value. Yahoo was eventually sold to Verizon in 2017 for approximately $4.5 billion.

AOL’s merger with Time Warner in January 2001 seemed to secure its survival through “real” media assets. Instead, the merger froze the company in place as broadband destroyed its dial-up cash cow. AOL’s $99 billion annual loss remains one of the largest in corporate history.

The false survivor pattern: interpreting survival as validation of the old model rather than license to build the new one. Cost-cutting disguised as strategy. M&A as substitute for innovation. Leadership focused on defending legacy revenue rather than cannibalizing it. During the quiet period of 2002-2005, Yahoo pursued search technology and advertising platforms through major acquisitions like Overture ($1.63B), while AOL contracted through layoffs and subscriber losses. In contrast, Google and Amazon invested heavily in engineering talent and infrastructure, with Google rapidly expanding its engineering workforce and Amazon increasing capital expenditures by 423%. The divergence wasn’t visible in stock prices – not yet. But it was visible in where the R&D dollars went.

Source: SEC filings (Google S-1 2004, Amazon 10-K 2005)

Some insurtechs currently surviving may prove to be false survivors – still losing money, growing gross written premium without improving combined ratios, burning furniture to stay warm rather than building better engines. The metrics to watch: loss ratio trajectory, customer acquisition cost trends, whether the company is building capabilities or just cutting costs.

The Legacy Carrier Risk

Traditional retailers felt vindicated after the dot-com crash. Borders, Barnes & Noble, and Sears concluded that “the internet is a catalog, not a store.” They halted digital innovation. The “Amazon is just for books” fallacy prevailed. By the time the iPhone launched in 2007 and mobile commerce began, these incumbents were structurally incapable of competing.

Newspapers celebrated the death of digital advertising competitors. They coined the phrase “trading analog dollars for digital dimes” as justification to delay digitization. When ad spend began tipping online in 2005, they ignored the shift – believing their “premium content” would protect them. Revenue freefall began in 2008.

The incumbent pattern: disruption threat triggers panic (1998-2000), crash triggers relief (2000-2001), false confidence leads to strategy paralysis (2002-2005), surviving disruptors emerge with superior unit economics (2006+), incumbent cannot catch up. The timeline from celebration to existential threat: 5-10 years.

Legacy carriers celebrating insurtech failures risk the same trap. The first wave of insurtech may struggle, but the technology—telematics/IoT sensors, AI underwriting, embedded distribution – is valid. Root’s telematics model is now showing profitability improvements after fixing base rates and expense structure. Coalition’s AI-driven risk assessment produces loss ratios carriers cannot match. Shift’s fraud detection is becoming industry standard rather than optional add-on. The survivors are not taking massive market share – not yet. They’re cherry-picking high-margin segments: cyber, small commercial, tech-forward auto. The segments carriers find operationally expensive to serve.

If carriers pause digital transformation now, believing the “disruption” is over, they will be blindsided by mature insurtechs – or tech giant entries – within 3-5 years. The quiet period is the most dangerous time to relax.

The Investment Implication

If the insurtech timeline follows the dot-com timeline: 2025-2027 should see continued consolidation with survivors clarifying. 2028-2030 should see clear insurtech winners emerge at scale. 2032-2035 should see full return to 2021 funding levels with sustainable models. 2035-2040 should see legacy carriers face existential pressure as digital-native models achieve dominance in profitable segments.

The public market recovery has already begun. Root has risen from $3.46 to approximately $74 – over 2,000% from trough. Lemonade has recovered from $10.27 to approximately $82. These are early signals, not endpoints. The dot-com survivors took 7-15 years to reclaim peak valuations and then exceed them by orders of magnitude.

Capital allocation in P&C insurtech has undergone a structural shift toward B2B technology vendors. In Q1 2025, 61.4% of all P&C insurtech deals went to B2B companies – a stark reversal from 2019-2021, when consumer-facing full-stack carriers and distribution models commanded the majority of venture capital. Traditional momentum investors, the “tourists” – Tiger Global, SoftBank Vision Fund, and other generalist VCs – have largely exited insurtech following significant portfolio losses and strategic pivots. The remaining active investors are predominantly specialists and strategic corporate venture arms: MS&AD Ventures, Brewer Lane Ventures, ManchesterStory Group, Aquiline Capital Partners, and select insurance incumbents making selective bets. Munich Re Ventures, once a leading CVC player with $1.2 billion deployed, is winding down operations by mid-2026, further consolidating the shift toward specialized insurtech-focused investors. They’re conducting deep diligence on loss ratios and distribution costs. They’re accepting longer timelines.

The pattern recognition is straightforward. The crash was real. The destruction was extensive. But the survivors are building during the quiet period, just as Amazon and Google built during 2002-2005. The incumbents celebrating too early are positioning themselves for the same fate as Borders and newspapers.

The next Amazon is currently building in the dark. The next Borders is currently celebrating its survival.

Author Note: This analysis draws on publicly available academic research, industry data, and regulatory filings. Statistics are cited to primary sources where available.

AI Disclosure: Research compilation utilized AI tools to discover and verify publicly available data sources and citations. All analysis, interpretation, and conclusions are original work.

Every company that you evaluate that is generating sensor data. Every construction equipment manufacturer that is retrofitting fleets with telematics. Every insurance carrier is sitting on petabytes of driving data, desperately trying to find a signal in the noise. They all face the same fundamental paradox: they are drowning in data, yet they are starving for value.

The global insurance market addressable by telematics data is projected to reach between $330 billion and $400 billion by 2033. It represents the single most mature dataset in the Internet of Things (IoT) ecosystem. Yet, according to Gartner, 80% of companies attempting to monetize IoT data will fail.

The IoT Paradox

The global insurance telematics market is projected to reach $400 billion. Yet, according to Gartner, 80% of companies attempting to monetize IoT data will fail.

These failures rarely stem from a lack of data availability. The data demonstrably transforms operations when applied correctly. Nor do they stem from technological incapacity; we possess the ability to process petabytes of streaming information in near real-time. Rather, these failures are structural. They occur because founders and investors fundamentally misunderstand the physics of data value creation.

After processing billions of miles of telematics data and observing the lifecycle of hundreds of companies – from unicorn exits to quiet bankruptcies – distinct patterns emerge. Whether a company is attempting to monetize vehicle data, industrial sensor readings, or agricultural IoT streams, the same principles determine the outcome. The telematics market offers two decades of expensive, hard-won lessons that every data-driven venture must understand to avoid becoming a statistic in the next cycle of capital destruction.

The Great Misdirection: Data Is Not the Product

The most pervasive fallacy in the IoT economy is the belief that data is the product. It is not.

Data is a raw material. It is unprocessed inventory – costly to store, unrefined, and useless in its natural state. It only accrues value when it is refined and transformed into specific outcomes for specific customers who are experiencing specific pain points.

To navigate this landscape, we must draw a sharp strategic line between Product Data and Data Market Data.

Product Data (The Internal Loop) is an operational utility. It is the dashboard a fleet manager uses to route trucks or the alert a foreman uses to schedule maintenance. It is a feature of the hardware, designed to improve the asset operator’s own P&L.

Data Market Data (The External Loop) is asset monetization. This is data extracted from the operation and sold to a third party – such as an insurer or traffic planner – to build a new product that is not developed by the raw telematics data supplier and does not exist within the fleet’s operation.

The fundamental error most ventures make is attempting to sell Product Data (raw reports and dashboards) to Data Market buyers. Insurers and hedge funds do not want reports; they want the signal to build their own derived products.

The “Product Data” vs. “Data Market” Framework

Watch how this confusion plays out identically across industries:

1. The Transportation “Dots on a Map” Fallacy

For years, we operated under the assumption that the product was GPS coordinates and speed metrics. We sold “dots on a map.”

But the flaw in this model was exposed by a common call to our sales team: “Thanks to your system, I identified the bad drivers and fired them. The rest are scared. I’m keeping the box in the truck as a dummy deterrent, but I’m cancelling the subscription.”

The business model collapsed because we solved a finite problem with a recurring fee.

To survive, the industry had to pivot to problems that never end. Fleet managers do not buy data. They do not want another dashboard.

They buy risk mitigation. They buy video telematics to exonerate drivers and avoid lawsuits. They buy safety tools to reduce accident frequency and lower insurance premiums. They buy the ability to keep their drivers on the road and their assets out of the courtroom.

They buy operational efficiency. They buy maintenance schedules that prevent costly roadside breakdowns. They buy engine diagnostics that predict failure before it happens. They buy routing that shaves miles off a delivery route. They buy electronic logging to meet hours of service requirements.

When fuel represents 25% of a fleet’s operating costs, even a modest data-driven improvement delivers substantial P&L impact. The data was always valuable, but the “product” – raw visibility – was not. By failing to translate raw signals into financial outcomes, early providers saw their value propositions erode into commoditization.

2. The Construction Utilization Gap

In the heavy equipment sector, manufacturers spent years installing sensors that tracked every hydraulic pressure variance and engine parameter. They offered this data to contractors, who largely ignored it. Why? Because a site foreman does not have the time to interpret hydraulic pressure charts.

However, when the industry reframed the data around utilization – specifically addressing the fact that heavy equipment utilization averages only 50%-70% – the market shifted. The product became “reducing idle time by 20%.” The data remained the same, but the translation of that data into an operational directive created a billion-dollar efficiency market.

3. The Agricultural Yield Revolution

John Deere does not sell soil moisture readings or GPS coordinates to farmers. If they did, they would be competing with free weather apps. Instead, they sell yield improvements.

By integrating sensor data into precision application machinery, they allow farmers to reduce fertilizer and pesticide inputs while increasing output. The result is a yield improvement up to 15%. The sensor data is merely the mechanism; the product is the margin expansion for the farmer – a reality John Deere now codifies with their ‘Application Savings Guarantee’, where fees are waived if the outcomes aren’t delivered.

The pattern is universal and unforgiving: successful data monetization requires deep domain expertise to transform raw signals into business outcomes. Technology companies frequently assume data has inherent value. Industry operators know that value only exists at the point of application.

The Three Eras of Every Data Market

Having observed the telematics sector evolve from satellite systems costing $3,000 per unit to ubiquitous smartphone applications, we can categorize the maturation of data markets into three distinct eras [1]. This framework allows investors to identify exactly where a portfolio company sits on the maturity curve.

The Three Eras of Data Maturity

Era 1: Infrastructure Without Business Models

In this pioneer phase, the focus is entirely on “plumbing.” Companies build the pipes to extract data, often without a clear idea of who will pay for it.

In 1996, General Motors launched OnStar. It was a technological marvel, yet GM struggled to monetize it beyond basic emergency services [2]. Similarly, Qualcomm’s OmniTRACS dominated the trucking sector through expensive satellite hardware [3]. Everyone was collecting data, but the infrastructure was designed for internal reporting, not external monetization.

Critically – and we see this mirrored in today’s emerging industrial IoT markets – the infrastructure of Era 1 is often incapable of exporting data. These systems are walled gardens, designed to ingest data but not to share it, creating massive technical debt that hinders future monetization.

Era 2: The Monetization Breakthrough (and Disruption)

This era begins when a market participant successfully sells data to a third party, proving that the “exhaust” of one industry can be the fuel for another.

In telematics, this breakthrough occurred in 2005 – years before the iPhone – when a major search engine executed one of the first commercial purchases of fleet GPS data for its map team [1]. This was revolutionary. Commercial fleet data was used to power real-time traffic algorithms for the general public. The value of the data was decoupled from the original hardware.

This era created massive value events. Nokia acquired NAVTEQ for $8.1 billion to secure map data ownership.

However, Era 2 inevitably invites disruption. In the mapping sector, Waze and the proliferation of smartphones destroyed the commercial fleet data model for consumer mapping within 36 months. Crowdsourced data from millions of consumer phones rendered commercial fleet data redundant for consumer mapping [1]. The lesson is stark: generic data applications always commoditize. Specific applications maintain value. Commercial vehicle data retained niche value for truck-specific routing (e.g., height and weight restrictions), but the mass-market value evaporated.

Era 3: Fragmentation Before Consolidation

We are currently living through Era 3. Instead of consolidating into a unified standard, the telematics data market has fragmented. Specialized solutions have emerged for every vertical.

Equipment manufacturers are building proprietary “walled gardens.” Other providers are building bespoke solutions. Today, 65% of large fleets utilize multiple, incompatible telematics systems. Whoever successfully integrates across these fragmented platforms will capture enormous value, not by generating new data, but by normalizing the chaos.

The Integration Crisis: Why Operations Beat Algorithms

The most significant barrier to data monetization in the insurance sector is not a lack of data; it is an inability to ingest it.

As detailed in the article I wrote for Carrier Management, Why Insurance Telematics Integrations Fail, insurance carriers are physically unable to utilize the billions of miles of telematics data they theoretically have access to. This is not a failure of intent; it is a failure of architecture.

The vast majority of legacy carriers operate on mainframe systems that date back to the 1970s and 1980s. These systems operate on a batch-processing logic. They are designed to process static forms once a year. Telematics providers, conversely, deliver real-time JSON streams containing millions of data points.

These two worlds do not speak the same language. They do not exist in the same time paradigm.

The Integration Clash

This integration failure is not unique to insurance. We see identical patterns in construction, where equipment tracking systems cannot talk to project management software. We see it in agriculture, where planting data sits in a silo separate from harvesting logistics. We see it in maritime shipping, where vessel telemetry is isolated from port authority systems.

The mundane, unglamorous challenge of making disparate systems communicate defeats brilliant algorithms and massive datasets every time. Companies that solve this integration problem capture disproportionate value. They do not win through technical superiority; they win through operational excellence and “boring” infrastructure.

The Unchanging Physics of Data Value

Across two decades of technology cycles- from satellite to cellular, from 3G to 5G, from simple GPS to computer vision – certain principles of data value have remained immutable.

1. Relevance Beats Volume

In 2009, the Washington State Department of Transportation (WSDOT) conducted a landmark study that fundamentally altered the trajectory of public sector data. The results proved a counterintuitive truth: high-frequency data from a small sample of vehicles provided superior insights compared to low-frequency data from a massive fleet. This wasn’t just a technical finding; it was the commercial proof point that opened governments at all levels to utilize telematics for land use planning. The lesson remains immutable: “Signal over Noise” is what buyers pay for. More data often simply equals higher storage costs; the right data, validated by a trusted authority, is what creates a market.

2. Specificity Prevents Commoditization

Generic data always trends toward zero marginal cost. Traffic data, once a premium product, was commoditized by crowdsourcing. Weather data is provided freely by NOAA.

However, specific, high-context data retains pricing power. While generic traffic flows are free, accident reconstruction data for insurance claims commands a premium. While regional weather is free, hyper-local soil moisture data for variable-rate irrigation is worth thousands of dollars per farm annually. The narrower and more specific the application, the more sustainable the economic moat.

3. Infrastructure Enables Everything

The invisible layer is where the actual war is won. It doesn’t matter if you have a breakthrough neural network if you cannot deliver the data into a legacy client’s workflow. The vast majority of failures occur because modern, real-time JSON streams crash against the walls of 1980s batch-processing mainframes. Companies that prioritize this “plumbing” – ensuring reliable delivery and lineage – build a competitive moat, while those relying solely on algorithmic superiority find themselves with a product that no one can ingest.

4. Operations Trump Technology

Throughout every era of the telematics evolution, companies with superior operations have defeated those with superior technology.

Consider the divergence between Progressive Insurance and the first wave of “Insurtech 1.0” companies. Progressive built a highly profitable, dominant business model using relatively basic telematics data. Conversely, heavily funded insurtechs, armed with advanced AI and superior user interfaces, lost billions of dollars.

The difference was not the technology; it was the operations. Progressive possessed the “scar tissue” required to manage claims, handle exceptions, and navigate regulatory environments. Technology without operational excellence is merely an expensive way to fail faster.

5. Trust is the License to Operate

Telematics data markets are about trust. Without trust, customers won’t give permission for you to sell the data. Without trust, no one is going to sell you their data. Companies have to trust you with their data.

Years ago, a leading telematics service provider sold their customer data and it was used by local law enforcement to set speed traps. That destroyed trust.

On the subject of trust, the old banking rule of “know thy customer” applies to anyone selling raw telematics data.

6. The Value Exchange Must Be Explicit

Customers inevitably ask: “What is in it for me?” It is their data; the technology company is merely the custodian. There must be a clear, tangible benefit for customers to grant permission for monetization.

This principle was crystallized at an industry conference where a Walmart executive stated they would be willing to share their proprietary logistics data – but only if it resulted in better roads and more efficient infrastructure usage. The trade was explicit: data for efficiency. Without that clear return on assets, the data remains locked.

7. The Prime Directive: Protect the Core

Whatever you do in the data business, you must always protect the core business that generates that data. This is the prime directive.

If your data monetization practices alienate your core customers, you will lose those customers. If you lose the customer, the data stream evaporates. You cannot build third-party products from a client base you have destroyed. There is no data market without a healthy product market to sustain it.

The Patient Capital Advantage

The realization that hardware and deep tech require long time horizons is now well understood. The market must now apply that same logic to data monetization.

The telematics market proves definitively that data maturity takes time. Between 2021 and 2024, the number of active investors making multiple deals in the insurtech space plummeted by 72%, dropping from 406 to just 113. This was not a cyclical dip; it was a structural correction.

Venture Capital funds discovered a fatal mismatch. Insurance transformation requires a 7-10 year horizon to accumulate sufficient data for actuarial credibility. You cannot rush the “law of large numbers.” However, VC funds operate on 10-year lifecycles, requiring exits within 5-7 years. The math simply does not work.

The Capital Timeline Mismatch

This capital timeline mismatch creates a fatal internal conflict. In the early stages of a data market, there is a fierce struggle for resources inside the producers of raw telematics data. The companies generate an enormous amount of data, but it is not central to their core business. Consequently, data teams struggle to get on engineering schedules to produce data in a form that can be used by third parties.

Timing matters. There must be enough buyers in the market to justify this internal investment. While some data markets, such as telematics and sensor data for traffic analytics, are large enough to support robust trading, others remain niche. Truck drivers need truck specific data – routes restricted by height or weight – but the market for this type of data is small. There simply are not enough buyers to support everyone who could sell this raw data.

If data markets don’t develop fast enough, management may pull the plug on what they consider to be a side business. Some data markets will simply not become large enough to justify the internal investment in supplying them. Some observers assume data guarantees profit. But innovators often build products nobody wants. This is an opportunity for patient capital.

This retreat has created an unprecedented opportunity for family offices and permanent capital structures. Unlike VCs, who are forced to manufacture growth to meet artificial fund timelines, family offices can align their capital with the reality of the industry. They can support companies through the necessary “Valley of Death” – the infrastructure building , the data accumulation, and the market education – to reach the scale achievement phase.

This patient capital thesis applies to every IoT data market:

• Industrial IoT: Platforms need 5-7 years to achieve critical mass and displace legacy systems.

• Smart Cities: Applications require multiple municipal budget cycles for adoption.

• Healthcare Data: Platforms face 3-5 year regulatory approval processes before monetization begins.

• Agriculture: Solutions need multiple growing seasons to prove ROI to skeptical operators.

The winners of the next decade will not necessarily be the companies with the most data or the most advanced algorithms. They will be the companies with capital structures that are aligned with the physical realities of their markets.

Actionable Advice for Data-Rich Ventures

If you are an operator sitting on sensor data – whether from robots, industrial equipment, or scientific instruments – the lessons from the telematics sector provide a clear playbook for monetization.

1. Find Your Outcome, Not Your Data Product

Stop organizing your business around selling data. Organize it around selling outcomes. Ask: What specific, measurable business result can we enable? Frame every product decision around that financial impact. If you cannot draw a straight line to a P&L improvement, you do not have a product.

2. Choose Depth Over Breadth

Resist platform ambitions in the early stages. The companies that try to be “the operating system for everything” invariably fail. The companies that solve specific, excruciating problems for specific industries succeed. It is far more profitable to own 80% of a niche vertical than 2% of a broad horizontal market.

3. Invest in the “Boring” Infrastructure

While competitors chase AI breakthroughs and press releases, invest in reliability, integration, and data governance. The unsexy foundations determine long-term success. The ability to integrate with a 1980s mainframe is a more valuable competitive advantage than a marginally better neural network.

4. Partner With Domain Experts

Technology alone never wins. You need industry insiders who understand the “hidden wiring” of the sector – the workflows, the regulations, the buying processes, and the politics. The marriage of technical capability and deep domain expertise creates true defensibility.

5. Structure for the Long Game

If your capital structure requires an exit in 3-5 years, you are already disadvantaged. Data markets require patience. Align your funding sources with your market’s maturation timeline. Do not take VC money for a Permanent Capital problem.

6. Stress-Test Your Data Viability

As a data market matures, producers of raw telematics or sensor data must ruthlessly determine the long-term viability of their asset before building the pipes to sell it. You need to map how pricing and market size will evolve.

The critical questions to ask are: Are there low-cost data substitutes? (If yes, margins will collapse). How specialized is the production? (If generic, it will commoditize). Will the number of buyers increase? (If you are selling to a market of one, you do not have a business; you have a hostage situation).

7. Respect the Supplier Saturation Point

If you produce raw telematics data, you must analyze the supplier capacity of the target market. Some data markets will grow and become robust, but they will structurally only need a limited number of suppliers to function. Once that capacity is filled, the window closes. If you miss the window, you must look for other markets.

Aside from your analysis, how are you supposed to determine what a market will look like? That is when you must remember the innovator’s reality: you face both the risk of product-market failure (no market demand) and the risk of poor timing (market saturation). You must distinguish between the two before you invest.

8. Navigate the “Shadow Phase” Strategically

In the early stages of a data market, real liquidity is silent. While the hype cycle produces press releases, the revenue cycle produces NDAs. This “Shadow Phase” is a feature, not a bug. Suppliers are testing the waters without alerting core customers to a monetization strategy that hasn’t yet proven its value. Suppliers require secrecy to test monetization without alarming their core customer base, while buyers demand it to hide their strategic roadmaps. If you interpret market silence as a lack of demand, you are misreading the room. By the time a data partnership is announced in a press release, you have already missed the early entry point.

9. Sequence Your Talent Stack Correctly

A common failure mode is hiring a Data Scientist when you actually need a Data Engineer. You must distinguish between the roles: Data Engineers build the pipes (moving data); Database Administrators (DBAs) maintain the storage (keeping the lights on); Data Scientists refine the product (extracting value). Do not hire a PhD Data Scientist to build your ETL pipelines. They will be frustrated, expensive, and ineffective. You cannot manufacture a product from inventory you cannot access. Build the logistics infrastructure (Engineers/DBAs) before you hire the alchemists (Scientists).

The Next $400 Billion

The telematics market reaching $400 billion is not the end of the story; it is the prologue.

Every sensor being deployed today, every IoT device being connected, and every data stream being generated is following the same evolutionary path that telematics blazed. Industrial IoT, smart cities, healthcare monitoring, and supply chain tracking are all in various stages of the same three eras.

The question is not whether these markets will mature – they will. The question is whether today’s innovators will learn from two decades of expensive lessons or whether they will pay the tuition again.

For those willing to study history, to understand market physics rather than just technological possibilities, and to execute with operational discipline, the opportunity is unprecedented. The data revolution isn’t coming; it is here. But value will not flow to those with the most raw inventory. It will flow to those who understand that data monetization is ultimately about transformation, not collection.

After 20 years of building in this space, one lesson stands above the rest: The winners aren’t the ones selling the data. The winners are the ones delivering the outcome.

What Comes Next

The laws of data physics we identified in telematics are not unique to transportation; they are universal. Yet, we see the same expensive architectural and business model errors currently being codified into the foundations of Industrial IoT, Smart Cities, and Digital Health.

This series continues by applying this operational lens to those emerging markets, distinguishing the inevitable winners from the infrastructure that will be rebuilt at a loss.

Endnotes

[1] Indenseo Research and industry analysis, 2025.

[2] Radius Payment Solutions. The History of Telematics. Radius Payment Solutions, 2023.

[3] Webfleet. The history of telematics: from the 1960s to today. Webfleet, 2023.

Author Note: This analysis draws on publicly available academic research, industry data, and regulatory filings. Statistics are cited to primary sources where available.

AI Disclosure: Research compilation utilized AI tools to discover and verify publicly available data sources and citations. All analysis, interpretation, and conclusions are original work.

The most telling signal in insurance today isn’t a demo; it’s a capital flow. Our research reveals the number of active investors repeatedly backing insurtech deals fell 72% in just three years – dropping from 406 active investors in 2021 to merely 113 in 2024 [1]. This isn’t a cyclical downturn or a “wobble.” It is a structural realignment. We’ve found this collapse stems from a fundamental mismatch: classic 10-year VC fund clocks simply do not match the 7–10 year build cycles required for insurance transformation [2].

The VC Exodus: A 72% Drop in Active Investors

The number of active investors repeatedly backing insurtech deals fell from 406 in 2021 to 113 in 2024

Source: CB Insights

The VC Exodus: A Failure of Capital Structure

Capital timelines broke product timelines. VCs discovered that even breakthrough technology can’t overcome the reality that insurance products need 3-5 years of actuarial data to achieve statistical credibility – a timeline that breaks their fund economics [2],[11]. This created a “valley of death” for scaling companies that had product-market fit but hadn’t yet proven loss-ratio impact.

When pitching the investment opportunity, a pattern emerged. Even supportive VCs admitted the capital environment was brutal. The market signal was clear: top-tier investors who had previously backed insurtech had soured on the sector after a string of disappointing returns. Counterintuitively, we found that VCs with prior insurtech investments weren’t a source of “smart money”; they were a source of “scar tissue.” They were less likely to take a meeting, not more.

These aren’t failures of innovation; they are failures of capital structure alignment

Consider the wreckage: Root Insurance’s valuation collapsed over 80% post-IPO [4]. Metromile, once valued at $1.3 billion, sold for a 61% loss [4]. Total insurtech losses between 2015-2020 exceeded $10 billion. These aren’t failures of innovation; they are failures of capital structure alignment.

The commercial auto insurance market crystallizes this problem.

The Commercial Auto Crisis: Key Metrics

The market has been unprofitable for 12 out of 13 years, with combined ratios consistently over 107%

Source: Insurance Journal, S&P Global

The line has been unprofitable for 12 out of 13 years [3]. Despite 55 consecutive quarters of rate hikes, combined ratios remain consistently above 107%, indicating a loss on every policy written [9]. Carriers still face nuclear verdicts averaging $23.8 million [3]. This is a market that requires patient capital to implement real solutions, but VCs, facing pressure for exits by year seven, have run out of patience and retreated.

The Patient Capital Revolution

Family offices are stepping into this void with a fundamentally different calculus. Managing between $5.5 and $10 trillion globally, they now provide 31-40% of global startup capital [2]. Unlike VCs facing fund expiration dates, they operate with permanent capital structures and multi-generational investment horizons. Our research identifies critical advantages they bring.

First, permanent capital and pacing. They can fund the 24-36 months of “data flywheel spin-up” and support companies through the complete 50-state regulatory approval cycle – a process that can cost $5-10 million before generating revenue [5].

Second, flexible value capture. They can capture value through multiple paths beyond a traditional exit: dividend distributions from profitable underwriting, partial strategic sales, and permanent hold strategies. The economics are compelling: by eliminating the 2-and-20 fee structure, they improve net returns by 300-500 basis points annually [5].

Third, and most importantly, operator-first governance. The dominant failure mode in insurtech was elegant technology without domain control. Family offices can enforce a “boring infrastructure first” agenda – focusing on policy admin, ingestion pipelines, and data governance. They can buy undervalued assets, layer in AI, and compound value by capturing loss-ratio improvement over years, not quarters.

The AI-Native Blueprint

The convergence of family office capital with artificial intelligence creates conditions we haven’t seen since the early internet era. AI has fundamentally altered insurance economics in ways that make family office investment timelines optimal.

Consider the transformation in underwriting costs: our analysis shows AI-powered systems can reduce processing from $1,090 per application to just $10 – a 109x improvement [5]. More dramatically, frontier AI training costs have collapsed by over 90% – with models like DeepSeek-V3 trained for $5.6 million, not $100 million – while inference costs have dropped over 280-fold since 2022 [10].

This catalyst shifts the entire competitive landscape. The infrastructure barriers that previously protected incumbents are crumbling. Our data indicates 70% of incumbent carriers’ digital transformation initiatives fail [6], primarily because they are attempting to modernize mainframe systems from the 1970s and 1980s [8].

With open-source models collapsing experimentation costs, advantage is migrating from “better algorithms” to better data, workflows, and distribution. The moat is now access to high-frequency, high-relevance data. Every telematics agreement, for example, needs explicit language on permissible uses, retention, and derivative analytics. AI also enables entirely new products: parametric coverage triggered by sensor data, dynamic pricing adjusted in real-time, and automated claims processing that reduces settlement time by 75% [7].

A Pragmatic Playbook for Patient Capital

For family offices ready to own this cycle, sequence matters. A pragmatic operator’s path involves six steps.

First, pick the wedge: Start where lift is measurable, like commercial auto programs with telemetry or specialty lines with dense operational data.

Second, acquire the pipes: Buy or partner with an integration shop that already aggregates multi-TSP feeds. Without the plumbing, AI is just theater.

Third, control the loss ratio, then the premium: Begin as an enablement layer (software + analytics). As the signal stabilizes, step into MGA authority, then add risk capital through a fronting/reinsurance stack.

Fourth, set capital-efficiency guardrails: Insist on the AI-native spend profile-milestone-gated tranches and shared savings, not blank checks.

Fifth, govern by outcomes: Track combined ratio trend, adjudication cycle time, and exoneration rate – not vanity ARR.

Finally, design for permanence: Build compliance, data governance, and model documentation from day one. It lowers audit friction and increases take-out value whether you sell, roll up, or hold.

Expect fewer loud brands and more “boring infrastructure” – quiet operators that own data rights, underwriting pipes, and claims muscle. Family offices that combine control deals with enablement platforms will compound value as the old stack finally gives way. Do the unglamorous work early and compound. That’s how durable moats form.

Endnotes

[1] CB Insights, State Of Insurtech Q4’23 Report, 2024. As cited in The Global Telematics & Sensor Data Market (Indenseo Research, 2025).

[2] Indenseo Research, Family Office Investment in Insurtech, 2025.

[3] Insurance Journal, Commercial Auto Combined Ratio Analysis, 2024; Conning, 2025 Commercial Auto Study, 2025.

[4] Fortune, “Root Insurance Post-IPO Performance Analysis,” 2023; TechCrunch, Metromile Acquisition Analysis, 2022.

[5] Indenseo Research, AI Impact on Insurance Operations, 2025; McKinsey & Company, InsurTech Capital Requirements Study, 2024.

[6] McKinsey Digital, Digital Transformation Failure Rate, 2024. As cited in The Global Telematics & Sensor Data Market (Indenseo Research, 2025).

[7] Accenture, AI in Insurance Underwriting, 2024; Deloitte, Insurance Innovation Report, 2024.

[8] Gartner, Insurance IT Budget Analysis, 2024.

[9] S&P Global, Commercial Auto Insurance Performance, 2025.

[10] Stanford University, 2025 AI Index Report, 2025; DeepSeek AI, DeepSeek-V3 Technical Report, 2024.

[11] Insurance Information Institute. Commercial Insurance Market Dynamics. III, 2024

Author Note: This analysis draws on publicly available academic research, industry data, and regulatory filings. Statistics are cited to primary sources where available.

AI Disclosure: Research compilation utilized AI tools to discover and verify publicly available data sources and citations. All analysis, interpretation, and conclusions are original work.