Why the World's Largest Carriers Can't Serve the Fastest-Growing Insurance Market

ETHAN:

That number is real. We pulled it from our analysis of 107 carriers operating in the U.S. market right now.

97% of them have minimum program sizes above $20 million. Their cost structures, their data infrastructure, their staffing models. All of it was designed for a specific shape of business: rare events, large policies, low frequency.

And that shape of business is not where the growth is.

The growth market is embedded insurance. $25 policies sold at the point of checkout. Youth sports registration. Gig worker coverage. SaaS platform add-ons. Hundreds of thousands of transactions per program per year.

The insurance market inverted. And almost nobody at the carrier layer noticed.

Today we're going to show you three things.

First, the data behind this inversion and why it matters for your business, whether you're running an MGA or exploring embedded insurance from a SaaS platform.

Second, why legacy carriers are structurally locked out of serving this market. Not because they're incompetent, but because their economic architecture won't let them.

Third, what carrier infrastructure looks like when it's built from scratch for the market that actually exists.

The first 45 minutes are education and original data. No product pitch. The last 10 minutes, I'll explain how we're building the carrier for the inverted market. If you want to leave at minute 45 with zero obligation, you'll still walk away with more actionable intelligence on embedded insurance than most people get in a year. Most people stay.

ETHAN:

Here's the plan for the next 55 minutes.

Section one: the market inversion. What shifted, why it matters, and what the data looks like. About 15 minutes.

Section two: the carrier gap. Why legacy carriers can't serve this market, with specific economics. About 10 minutes.

Section three: what purpose-built carrier infrastructure looks like. Architecture, not marketing. About 10 minutes.

Then we'll transition to what we're building at Slingshot and how to engage. About 10 minutes. And we'll close with Q&A.

I want this to be useful regardless of whether you ever talk to us again. So the education is real. The data is original. And the questions at the end are open to anything.

Let's go.

ETHAN:

Insurance has operated on the same fundamental economic assumption for 200 years.

The assumption is this: insurance covers rare, catastrophic events. Your house burns down. Your car is totaled. Someone sues your company for $10 million. These events are infrequent but severe. And every piece of carrier infrastructure --- from the actuarial models to the data pipelines to the staffing ratios to the regulatory frameworks --- was built around this assumption.

It worked. For two centuries, it worked brilliantly.

[Click to transition slide]

Here's what changed.

Insurance moved into the checkout flow. You register your kid for youth soccer and there's a $25 accident policy at checkout. You book a concert ticket and there's event cancellation insurance. You subscribe to a SaaS platform and there's embedded coverage for your business operations.

These policies are small. $25 on average. But there are hundreds of thousands of them per program per year. And the market is growing at 30%+ annually.

[Click to market sizing slide]

This is the embedded insurance market. And the carriers didn't notice. Or they noticed and couldn't do anything about it. Because here's the thing.

A legacy fronting carrier needs roughly $20 million in gross written premium per program to make the economics work. That's not greed. That's arithmetic. Their operational cost structure --- compliance staff, underwriting teams, reporting infrastructure, regulatory filings --- was built for large programs with relatively few policies. When you try to run a $3 million program through that infrastructure, the carrier loses money. So they don't do it.

State National's minimum is $20 million. Fortegra. Markel. Concert. All the same. One carrier literally laughed when a SaaS founder said her first-year book would be $3 million.

But that $3 million book, if it's 120,000 policies at $25 each, might grow to $10 million in year two and $30 million in year three. The growth is there. The carrier infrastructure isn't.

ETHAN:

This is the part most people get wrong. They think legacy carriers just need better technology. Bolt on some AI. Hire a CTO. Build an API.

It doesn't work. And I can prove it.

Swiss Re is a $35 billion reinsurance company. They built iptiQ, their own carrier-as-a-service platform, with essentially unlimited capital. They spent years and hundreds of millions of dollars.

They're selling it. They couldn't make it work.

Not because they didn't have the money. Not because they didn't have the talent. Because they built it on top of existing carrier economics. The architecture was wrong.

When you bolt AI onto a carrier system that was designed in the 1990s, the AI can only be as good as the data it touches. And the data at most carriers is a mess.

I know this from the inside. At Palomar, I was Chief Actuary and Chief Data Officer. I had access to everything. And I still couldn't trace the lineage of a single data object from origination to final report without calling three people and checking two systems.

That's not a bug. That's the architecture. Carriers don't have data lineage because they were never designed for it.

[Click to Snapshot vs. Restated slide]

Here's a concept most people in this audience have never heard of, because carriers don't talk about it.

When your carrier sends you a bordereau, is that number the snapshot --- the number as it was at the moment of calculation --- or is it the restated number, adjusted after the fact? Most carriers don't distinguish between the two. Numbers get quietly adjusted. You don't know which version you're looking at.

If that sounds abstract, ask Marcus. He had $3 million in collateral frozen for 18 months because his reinsurer asked one question: "Where did this premium number come from?" Nobody at the carrier could answer it.

ETHAN:

Let me put real numbers on this.

If you're running a $20 million MGA program with a legacy carrier, here's what the data architecture problem actually costs you.

So here's the question. What would carrier infrastructure look like if you built it from scratch? Not bolted AI onto a 1990s system. Not tried to modernize a legacy architecture. Built it new.

ETHAN:

This is where I get excited, so bear with me.

Imagine every single data object in a carrier system --- every policy, every premium payment, every claim, every bordereau entry --- carries what we call a birth certificate.

Not a timestamp. A birth certificate. Origin: where did this data come from? Chain of custody: who touched it, when, and what did they change? Version history: every state this data has ever been in, with the ability to see any prior version.

In this system, when a reinsurer asks "where did this number come from?" the answer takes 30 seconds. Not 30 days. Not "let me get back to you." 30 seconds. Because the lineage is baked into the data model, not added as an afterthought.

Reconciliation becomes automatic. Not a quarterly sprint. When data ingests, it reconciles against the source. Discrepancies flag in real time. The three-week, five-person reconciliation exercise becomes a dashboard.

Now layer on what we call the Institutional Memory Engine.

When an underwriter leaves your carrier, what happens to everything they knew about your program? It walks out the door with them. Their replacement starts from zero.

What if the platform retained that intelligence? Every underwriting decision, every claims pattern, every risk assessment, every exception --- stored as structured knowledge that persists regardless of who's sitting in the chair.

Day 1 performance equals Year 5 knowledge. Because the system learns from every transaction, across every program, from the moment it goes live.

And here's the part that matters for the inverted market.

Legacy carrier operations cost somewhere between $12 and $18 per policy to process. That's fine when the average policy premium is $5,000. The cost is invisible.

But when the average policy premium is $25? The processing cost is higher than the policy itself. That's why legacy carriers need $20 million minimums. Their cost-per-policy makes small programs uneconomic.

Purpose-built AI-native operations bring that cost below a dollar per policy. That's what makes a $3 million program profitable. That's what makes sub-$20 million economics viable. And that's what makes the inverted market accessible.

ETHAN:

Everything I've shown you so far --- the inversion data, the carrier gap analysis, the architecture concepts --- that's the problem and the framework.

Now I want to tell you what we're doing about it.

But first, one thing. If this is where you want to get off, you've already gotten what we promised: original market data, a framework for evaluating carrier partners, and three questions you should be asking any carrier you talk to. That's real. You can leave now and use this.

For those of you who want to see the solution, let me introduce Slingshot properly.

We didn't start with a carrier and try to modernize it. We didn't bolt AI onto a legacy system. We didn't license a platform and call ourselves a carrier.

We acquired a carrier shell with 50-state licensing and statutory surplus. And we're rebuilding the entire operational and data architecture from the first line of code.

Why? Because I spent 15 years inside carrier operations and I know that you can't fix the architecture from inside. The institutional resistance isn't about people. It's about economics. Every dollar of modernization cannibalizes the existing business. Nobody at a profitable legacy carrier will sign off on that.

So we built a new one.

ETHAN:

Here's how Slingshot works. Three tiers, each built for a specific type of partner.

ETHAN:

Before we open Q&A, let me address the five things I know you're thinking.