A Restaurant That Adds Happy Hour Three Nights a Week Is a Different Risk Than the One You Underwrote

Bobby Touran didn’t come from insurance. He studied business and economics at McGill, cut his teeth in early-stage investing and an incubator role starting around 2008-2009, and moved to the San Francisco Bay Area in 2017 to join HVF Labs, an incubator founded by Max Levchin (PayPal co-founder, later Affirm CEO) built around a specific model: bring in experienced founders as entrepreneurs-in-residence and let them find a problem worth solving.

Commercial insurance appealed to Bobby precisely because he knew almost nothing about it. Early conversations with people at Chubb introduced him to the excess and surplus (E&S) lines market and Lloyd’s of London, and he built his first company, Pathpoint (originally called Outline), as a streamlined digital wholesale broker for small commercial E&S risk — a quote-bind-issue platform giving retail agents direct API access to carriers like Crum & Forster, Chubb, Hiscox, and Beazley, cutting a process that traditionally took days or weeks down to minutes.

Five years into running Pathpoint (still operating today), Bobby started Rainbow — an MGA built on a different insight: instead of distributing other carriers’ products in the non-admitted market, build a specialized, software-driven underwriting operation in the admitted market, focused on a single vertical he knows and loves personally: restaurants.

In Episode 109 of InsurTechTalk, Bobby and I covered the mechanics of E&S versus admitted insurance, why restaurant risk changes underneath a policy without anyone telling the carrier, and what he learned watching a wave of earlier insurtechs struggle.

About Bobby Touran

Bobby Touran is the co-founder of Rainbow, an admitted-market MGA specializing in restaurant insurance for the small commercial segment. He previously founded and ran Pathpoint, a wholesale insurance distribution platform for small commercial excess and surplus lines business, which he led for roughly five years starting in 2017. Rainbow is reinsured through Accelerant and, as of this recording, was live in five states (Arizona, Texas, Indiana, Michigan, Utah) with a licensed footprint across 26.

Admitted vs. Non-Admitted, Explained Plainly

Bobby’s framing of the insurance landscape, given for listeners unfamiliar with the structure: P&C splits into personal and commercial lines; commercial splits further into small, middle-market, and large risk. Within that:

• Admitted policies are backed by state guarantee funds — if the carrier becomes insolvent, the policyholder retains some protection. Because it’s heavily regulated at the state level, selling an admitted product nationally requires separate approval from all 50 state departments of insurance
• Non-admitted (excess & surplus, or E&S) policies operate in a far less regulated, more free-market structure, without the state guarantee fund backstop — typically used for higher-risk or harder-to-place risk (his example: a restaurant in a hurricane-exposed part of Miami versus an identical restaurant inland)

Pathpoint operated in the non-admitted market as a distributor of other carriers’ products. Rainbow operates in the admitted market, underwriting its own specialized book — a meaningfully different regulatory and operational challenge.

Why Small Commercial Is a Genuinely Attractive, Fragmented Market

Bobby’s read on the small commercial P&C opportunity, developed through years working with retail agents at Pathpoint:

• It’s a $130-140 billion annual premium market
• It’s genuinely fragmented — even the largest players hold roughly 5% market share
• It’s historically been profitable for carriers and notably sticky — small business owners don’t want to re-shop mandatory coverage annually; they want to set it and move on

That stickiness, combined with fragmentation, is what made him believe there was room for a more specialized, technology-native underwriter rather than another broad-appetite generalist.

Why Restaurants

Within small commercial, Bobby zeroed in on food and beverage for a mix of strategic and personal reasons. Traditional carriers have pulled back meaningfully from hospitality risk over recent years, accelerated by COVID — creating exactly the kind of underserved gap Rainbow could target. It’s also, candidly, a category Bobby is emotionally drawn to; he described genuinely loving cooking and food as a personal interest that made the sector appealing beyond pure market logic.

The Restaurant Underwriting Problem: Risk That Changes Without Telling You

This was the sharpest technical insight in the conversation, and it’s specific to why restaurants are genuinely hard to underwrite well.

Carriers routinely underwrite a restaurant based on stated operating details — closes at 9pm, no happy hour, no live entertainment — and then discover at renewal that the actual business has drifted substantially: now open late, running happy hour daily, hosting live performances multiple nights a week. Bobby’s explanation of why this matters isn’t cosmetic — it directly changes the liability profile:

• Alcohol-driven happy hour promotions carry materially different liability exposure than a lunch-focused sandwich shop
• Extended operating hours simply increase the surface area for any kind of incident to occur
• A carrier that priced and bound a low-key neighborhood lunch spot is, by renewal, actually covering something closer to a late-night bar with entertainment — a different risk entirely, priced wrong from day one of the drift

Rainbow’s answer is continuous underwriting — not just careful assessment at binding, but ongoing monitoring of publicly available information (advertised hours, promotional content, reviews, imagery) throughout the life of the policy, specifically to catch this kind of drift early and open a dialogue with the agent before it becomes a renewal surprise or a claims problem.

Going Live: The Regulatory Reality of Admitted Business

Rainbow’s launch illustrates exactly the friction Bobby described as typical for any company building an admitted product from scratch. The company holds MGA licensing across all 50 states, but the restaurant program itself required carrier and reinsurer appetite plus individual state regulatory approval — a process Bobby candidly called “somewhat opaque” and slower than he’d like as a company trying to iterate quickly. Rainbow wrote its first policy in early August (having gone live in Arizona shortly before), with Texas following soon after, and a plan to add roughly 10 more states over the following six to eight weeks.

Agent Distribution Was a Deliberate, Learned Choice

Rainbow sells 100% through agents — a decision Bobby made having watched newer insurance entrants attempt direct-to-consumer distribution in small commercial and largely fail or pivot away from it. His reasoning: roughly 80% of small commercial insurance in the US is still sold through agents, and agents serve a genuine underwriting-quality function, not just a distribution one. A pizza shop owner is an expert in pizza, not in matching coverage to their actual risk profile — an owner shopping purely on speed and price, without an agent’s guidance, is far more likely to end up underinsured for something specific to their business (his example: unique equipment central to their value proposition), producing exactly the adverse claims experience that damages a portfolio’s loss ratio over time.

The Claims Loop: Keeping the Agent in the Conversation

Rainbow’s platform gives policyholders a dashboard to access their policy, message their agent directly (with the message surfacing on the agent’s own dashboard), and file a first notice of loss with a few clicks, kicking off the claims process with Rainbow’s claims partner. Rainbow’s underwriting and service team monitors claim status and keeps the agent informed throughout — preserving the agent’s role as the trusted relationship-holder rather than disintermediating them once a claim is filed.

What Earlier Insurtechs Got Wrong

I asked Bobby directly, having watched a wave of well-funded insurtechs (public and private) struggle or disappear, what he thought went wrong — acknowledging upfront that most of what he knows about competitors is necessarily limited and partly speculative.

His diagnosis, in three parts:

• Growth prioritized over sustainable, profitable growth. Many earlier insurtechs, in his assessment, simply weren’t focused on profitability from day one — a structural choice, not a market accident
• Manual operations disguised by cheap capital. In a favorable financing environment, it was often easier to throw human labor at a problem than build genuine software automation. Those costs compound over time, and when the financing environment tightens, companies are forced into layoffs that damage the momentum a startup depends on to survive
• Lack of specialization. Broad-appetite underwriting across dozens of business types makes it hard to build genuinely differentiated, industry-specific data infrastructure — maintaining data connections and expertise across many verticals becomes expensive and unwieldy in a way that undermines the loss ratio discipline needed to survive

His pointed critique of a common “innovation” claim in the category: pre-filling property details from a generic hazard data provider isn’t really underwriting differentiation — it’s a baseline feature, not a moat. Rainbow’s bet is that genuine vertical specialization (restaurant-specific data, restaurant-specific underwriting judgment) is what actually produces defensible loss ratio performance, in a way that broad, shallow appetite cannot.

Investor Fit Matters as Much as Capital

Bobby’s broader observation on the category’s public struggles: insurance and insurtech aren’t currently the most fashionable categories for outside capital compared to AI or crypto, which he considers, if anything, an advantage — it selects for investors who genuinely understand the industry’s underwriting economics rather than chasing hype-driven growth metrics disconnected from loss ratio discipline.

Advice: Give Software a Chance, and Take the Industry Seriously

Asked for closing advice, Bobby split his answer into two audiences. To insiders: don’t write off insurtech because some early, well-known companies struggled publicly — software is not leaving insurance, and the honest question worth asking is how technology can genuinely improve both your own operations and outcomes for your clients. To outsiders considering the industry: insurance is a genuinely remarkable field, built on relationships and multi-generational careers, and it underpins essentially everything functioning in modern life — reason enough to take it seriously rather than dismiss it as boring or legacy.

Key Takeaways

• Admitted and non-admitted (E&S) insurance are structurally different markets — regulatory backing, state-by-state approval requirements, and typical risk profile all differ meaningfully, and building a company in one doesn’t directly transfer expertise to the other
• Restaurant risk changes materially and often invisibly between binding and renewal — extended hours, added alcohol service, and live entertainment each shift liability exposure in ways a static annual underwriting cycle misses
• Continuous underwriting — monitoring public business information throughout the policy term, not just at binding — is Rainbow’s structural answer to that drift problem
• Agent distribution remains dominant in small commercial (roughly 80% of the market) because agents provide a genuine underwriting-quality function, not just a sales channel — direct-to-consumer has largely failed for newer small commercial entrants for this reason
• Vertical specialization in underwriting, not broad appetite, is what Bobby believes actually produces defensible, sustainable loss ratios — generic data prefill is a baseline feature, not real differentiation
• Prioritizing growth over profitability, and using cheap capital to paper over manual operations instead of building real automation, are recurring, structural reasons earlier insurtechs struggled
• Investor unfamiliarity with insurance economics is a real risk for the category, making genuinely insurance-literate capital more valuable than capital chasing broader tech hype cycles