Why Isn’t Anyone Asking Startups What Loss Ratio They Targeted, and What They Actually Hit?

Joe Emison, co-founder and CTO of Branch Insurance, came into this conversation with a specific agenda: he’d been listening to the broader fintech and insurtech commentary landscape and concluded there’s a real, correctable gap in how thoughtfully US personal lines insurance — home, auto, renters, and to some extent life — actually gets discussed. What followed is one of the most analytically dense conversations in the show’s history, working through insurance economics from first principles rather than borrowed software-as-a-service intuition.

In Episode 54 of InsurTechTalk, Joe and I covered whether home and auto insurance are genuinely commodities, the three separate buckets of customer acquisition cost that most analysts and S-1 filings never fully account for, the cautionary history of Esurance, and why every major personal lines carrier in US history got big the exact same way.

About Joe Emison

Joe Emison is co-founder and CTO of Branch Insurance, a home, auto, renters, and umbrella insurer headquartered in Ohio, operating as an MGA in some states and as its own carrier — specifically structured as a reciprocal exchange — in others. At the time of this recording, Branch was live in 14 states with plans to reach all 50 plus DC, had posted two consecutive years of more than 10x growth, and had recently closed a $50 million Series B led by Anthemis.

Are Home and Auto Insurance Commodities?

Joe framed the entire conversation around one foundational question every insurtech founder and analyst should answer explicitly: are personal lines products like home, auto, and renters insurance genuinely commodities that people buy on price above a basic trust threshold, or do consumers actually perceive meaningful differences between providers? His own answer, for the middle of the market (excluding very high-value homes or specialty vehicles where replacement cost genuinely gets complicated): these products are overwhelmingly commoditized, and people shop almost exclusively on price and ease of purchase.

He backed this with concrete evidence from Branch’s own early product research: an open-ended “magic wand” survey asking people what they’d change about their last home or auto insurance purchase, with no prompted options, found 85% of unprompted responses centered on wanting it to be cheaper. He also shared a personal anecdote from Branch’s early days — being genuinely surprised listening to sales calls when a customer became visibly excited about saving just $20 over a six-month policy period, evidence of just how price-sensitive buyers really are. He cited a line he’s heard repeated in the industry: if you’re selling underpriced auto insurance, people will find you even if you’re sitting in a dinghy in the middle of the ocean.

Genuine Coverage Differences Are Rarer Than People Think

I pushed on whether coverage differences matter more for a smaller, more sophisticated segment of “confident planner” buyers. Joe’s response was direct: across mainstream US personal lines, it’s genuinely difficult to find meaningful coverage that one major carrier offers that another doesn’t (he named sewer and water line coverage on the homeowners side as one of the few real exceptions). He drew a useful conceptual line for what counts as real insurance versus something else entirely: insurance exists to protect against catastrophic financial loss, and as claim dollar amounts get smaller, a product starts feeling less like insurance and more like a subscription. His example: Plymouth Rock offers an endorsement bundling Red Sox tickets and memorabilia — a fun perk, but not insurance in any meaningful sense.

His broader point: it’s genuinely hard to find a US personal lines carrier missing a coverage that matters to even 1% of the population. He was direct that anyone claiming otherwise likely doesn’t fully understand how mature and comprehensive the mainstream personal lines market already is — you can call GEICO, Progressive, Allstate, or USAA directly, talk to a licensed agent, and walk away confident you have the right coverage, regardless of which carrier you chose.

The Real Problem: People Don’t Know What They Bought

Where Joe agreed there’s genuine dysfunction: most people’s actual experience with home and auto insurance is buying it because they’re required to, then only discovering during a claim whether they were actually covered for what happened. At Branch, he described this outcome — a customer calling in with a claim they believed was covered, only to learn it wasn’t — as a defect the company actively works to prevent, largely through real human service agents rather than leaning entirely on AI and chatbots, since Branch doesn’t believe that approach solves the underlying comprehension problem. He agreed there’s a real difficulty in helping consumers compare coverage apples-to-apples across carriers, but maintained that within any single carrier relationship, especially one with knowledgeable licensed agents, consumers who take the time to ask questions generally do end up with the coverage they actually need.

The Three Buckets of Customer Acquisition Cost Nobody Fully Counts

This was the sharpest, most original framework in the conversation. Joe argued that buying home and auto insurance in the US remains genuinely difficult — a full home-auto bundle purchase can take hours and hundreds of questions, and even a carrier as innovation-forward as Progressive requires roughly 15 minutes for a non-bindable auto quote, another 15-20 minutes for a non-bindable home quote (often actually placed with a different carrier Progressive represents as an agent, like ASI, QBE, or Homesite), followed by separate purchase steps and, in some cases, an additional call to actually bind home coverage.

That friction is exactly why insurers spend so heavily to acquire customers — and Joe argued most analysts and S-1 filings miss two of the three real buckets of that spend entirely:

• Sales and marketing spend — the bucket everyone already counts (Allstate’s “Mayhem” and hands campaigns, State Farm’s “Jake from State Farm,” and similar brand-building spend aimed at top-of-mind recall)
• Agent commissions — often around 15% of premium, and critically, this commission typically persists for the entire life of the policy, not just the initial sale. Joe’s framing: if an agent isn’t providing ongoing value commensurate with that recurring 15%, the agent is getting a better deal out of the relationship than the customer or the carrier is, since the agent’s core economic role is acquiring and binding the customer, not necessarily servicing them afterward (a role carriers often retain directly regardless)
• Underpricing disguised as growth — any startup running a loss ratio meaningfully above its actual target (and certainly above 100%, where claims paid out exceed premium collected) is effectively subsidizing growth through mispriced insurance, which functions as a disguised, unsustainable acquisition expense, since underpriced insurance is exactly the kind of growth people are more than happy to buy

His conclusion: the reason GEICO, Progressive, and USAA have been the fastest-growing US auto insurers over the past two decades comes down almost entirely to not paying that recurring agency commission — literally the origin of GEICO’s “15 minutes could save you 15% or more” pitch, since that 15% is roughly what a traditional agency-distributed competitor has to build into its price that GEICO simply doesn’t.

A Borrowed Framework: LTV and CAC, Adapted for Insurance

Joe pointed to investment bank William Blair’s published work adapting software-as-a-service metrics for insurance economics as a genuinely useful independent reference — their approach treats an insurtech’s effective LTV as premium collected minus the cost of claims (essentially one minus the loss ratio), letting investors compare insurtechs against established carriers using consistent, comparable investor-facing metrics. He recommended it directly to any investor active in the space who hadn’t read it yet.

Why Combined Ratio Is the Wrong Metric for a Startup, and Loss Ratio Gaps Are the Right Question

Joe was direct that combined ratio is a poor lens for evaluating an early-stage insurtech, since it’s fundamentally a scale game and any young company’s combined ratio will look poor simply because it hasn’t reached scale yet. The metric worth focusing on instead is loss ratio specifically, and more precisely, the gap between a company’s targeted loss ratio and what it actually achieved. Setting a loss ratio target is fundamentally a forward-looking guess about the true cost of a product — not just claims from crashes or fires, but also the real, often-overlooked servicing costs (billing method changes, failed payments requiring certified mail and follow-up calls, and similar operational friction) that ultimately get spread across the existing customer base, since an insurance company isn’t the government and doesn’t have an unlimited pool of money to absorb those costs.

His pointed observation: he’s never once heard an analyst directly ask an insurtech the simple, revealing question — you targeted this loss ratio, you hit this loss ratio, why the gap? — despite how much that single question would reveal about whether a company genuinely understands how to underwrite its own product.

The Esurance Cautionary Tale

Asked to illustrate what happens when that discipline is missing, Joe told the history of Esurance, the original internet-native auto insurance startup, built on the simple thesis that people would want to buy insurance online. Esurance copied Allstate’s rating plan — legal, since rate plans must be filed and are publicly available — but rating plans don’t capture underwriting rules, the separate, unfiled criteria determining who a carrier is actually willing to offer a given price to. Allstate’s program was designed as a preferred-tier product, restricted to lower-risk drivers; Esurance copied the pricing without replicating that underwriting discipline, then marketed the resulting product broadly and aggressively online. The result, per Joe’s account: the company attracted exactly the population the original pricing was never designed to cover, structurally undermining the business almost from the start. Esurance was eventually sold to Allstate for roughly a billion dollars, but Joe’s understanding is the original founders made comparatively little from the outcome given how much the underlying underwriting mismatch had already damaged the business.

The Counterintuitive Truth About Loss Ratio Targets

Joe offered a genuinely counterintuitive insight worth sitting with: a company targeting a higher loss ratio than a competitor, using a comparable rating plan, is actually the more efficient business, not the less disciplined one. A targeted loss ratio effectively defines how many cents of every premium dollar remain for everything else — sales and marketing, agent commissions, profit. A company that only needs 25 cents of every dollar for those other costs (supporting a 75% loss ratio target) can charge less than a competitor that needs 40 cents (a 60% loss ratio target), for the same underlying risk — meaning the company with the higher loss ratio target is structurally cheaper and will convert better in the market, all else equal. His challenge to any startup targeting a notably lower loss ratio than efficient, mature incumbents like Progressive or GEICO: if you don’t carry the overhead of legacy mainframes and decades of accumulated infrastructure, why do you need more margin than they do, not less?

He was careful to note a legitimate exception: a startup can reasonably separate its true at-scale loss-and-expense economics from temporary, investment-funded brand-building or growth capital expenditure, as long as those buckets are clearly and honestly disclosed separately rather than blended into a headline loss ratio number that obscures the real underlying unit economics.

Why SaaS Investor Instincts Actively Mislead in Insurance

We discussed directly why generalist, SaaS-background investors can be genuinely dangerous evaluators of insurance startups. In SaaS, once you’ve paid to acquire a customer, that customer is typically profitable for as long as they remain a customer — the entire investment thesis rewards prioritizing growth now and monetization later. In insurance, Joe was blunt: if you’ve mispriced the product, every single minute that customer remains yours, you’re losing money — a leaky-bucket dynamic with no equivalent in software economics. Worse, insurance regulation generally prevents a carrier from simply dropping a customer because the company itself misjudged its own pricing, since doing so would be genuinely disruptive to that policyholder’s life. His conclusion, backed by hard historical example: nobody has ever successfully executed a strategy of deliberately underpricing insurance to grow fast and “fixing it later.”

How Every Major Carrier Actually Got Big

Joe made this point with genuine conviction: State Farm, Farmers, Allstate, Travelers, USAA, and Progressive all got large the same fundamental way — by having a structurally cheaper price than the rest of the market for a sustained period, and passing those savings through to customers.

• State Farm was founded by a farmer who recognized that auto insurers were charging rural drivers the same as urban drivers despite meaningfully lower rural accident exposure — when incumbents dismissed the idea, he built his own company charging farmers less, and it worked
• Farmers copied State Farm’s basic model but innovated on commission structure specifically, paying agents commission only in the policy’s first year rather than for its full life — enabling a structurally lower ongoing price that, despite the lower per-agent payout, still drove far higher volume
• Allstate leveraged its position inside the Sears catalog to acquire customers at meaningfully lower cost than competitors relying on traditional agency distribution

His summary: every major personal lines success story in US history reduces to the same underlying mechanism — a structurally cheaper product, sustained over time, with savings passed directly to the customer. Anyone building an insurtech today should be able to explain clearly whether their own growth thesis follows that same historical pattern, or genuinely departs from it and why.

Branch: Cheaper Because It’s Easier to Buy

Turning to Branch itself, Joe described its core strategy as removing acquisition cost from the purchase process by making Branch dramatically easier to buy than the traditional alternative. Branch is a home, auto, renters, and umbrella insurer headquartered in Ohio — its first state, followed by Texas and Arizona as its second and fifth states — live in 14 states at the time of recording, with plans to reach all 50 states plus DC. It operates as an MGA in newer states and migrates to its own paper (structured specifically as a reciprocal exchange, working initially through a reinsurer’s paper before transitioning) in states where that structural transition has had time to complete, since operating on its own paper carries real pricing advantages. He contrasted this with Clearcover’s strategy of deliberately launching first as an MGA in California — one of the hardest possible states to start in, on the theory that every other state becomes comparatively easier afterward.

Branch had posted two consecutive years of more than 10x growth at the time of recording, and had closed a $50 million Series B led by Anthemis earlier that year, with Joe expecting to raise again in 2022.

Advice: Go See Twenty One Pilots Live

Asked for a closing recommendation, Joe pointed to Twenty One Pilots, a band that grew up and found fame in Columbus, Ohio, having recently taken his twelve-year-old twins to see them live. He described their live performances as genuinely genre-bending — moving from hardcore rap territory into a cover of “Benny and the Jets” within the same set — and specifically compelling to experience live rather than just on record. His broader closing thought: more people should make a point of getting out to live music and live theater generally.

Key Takeaways

• Home, auto, and renters insurance are overwhelmingly commoditized products for the middle of the market — Branch’s own unprompted “magic wand” customer survey found 85% of respondents simply wanted it to be cheaper, with no prompting toward price as an option
• True customer acquisition cost in insurance spans three distinct buckets most analysis only partially captures: conventional sales and marketing spend, recurring agent commissions that persist for the life of a policy, and underpricing that functions as disguised, unsustainable growth subsidy
• The most revealing diagnostic question for any insurance startup is the gap between its targeted loss ratio and its actual loss ratio — a question Joe says almost never gets asked directly, despite how much it reveals about underwriting discipline
• Esurance’s history is a cautionary tale of copying a competitor’s public rate filing without replicating its unfiled underwriting rules, attracting exactly the risk segment the pricing was never designed to cover
• Counterintuitively, a company targeting a higher loss ratio than a competitor (using a comparable rating plan) is the more efficient, structurally cheaper business — since the loss ratio target defines how much margin remains for everything else, and needing less of that margin means a lower, more competitive price
• SaaS-style investor instincts (accept losses now, monetize later) are actively dangerous applied to insurance, since a mispriced policy loses money every day it remains active, and regulation generally prevents simply dropping underpriced customers to fix the mistake
• Every major US personal lines carrier — State Farm, Farmers, Allstate, and others — grew large through the same underlying mechanism: a structurally cheaper price sustained over time, with savings passed directly to customers, not through short-term subsidized growth