Capital, Underwriting, Distribution, Trust: A VC’s Framework for Where Insurance Value Actually Lives

Noam Inbar’s path into venture capital started with a red card.

She spent her teenage years as a soccer referee in Israel — youth leagues at first, then working her way up through a career that got “delayed,” in her words, by a few incidents that reflect just how aggressive Israeli soccer can get. She kept refereeing all the way through her military service. It is not, on paper, a typical origin story for a General Partner at one of Israel’s most prominent venture capital groups.

But by the time Noam sat down with me for Episode 168 of InsurTechTalk, that soccer-referee instinct — reading a room, managing conflict, making a call under pressure with incomplete information — had clearly translated into something else: a genuinely sharp framework for where value actually sits in the insurance industry, and why AI is rearranging that value in ways most investors haven’t fully priced in yet.

About Noam Inbar

Noam Inbar is General Partner at Viola FinTech, part of Viola Group, the largest technology investment group in Israel. Before venture capital, she built a career as an operator across some of the most consequential fintech companies of the last two decades: an internal marketplace venture at Orange (Israel’s largest telco) around 2007–2008, before e-commerce existed in Israel in any real form; PayPal’s first Israel office, where she discovered her love of startups and fintech before either was considered cool; Zooz, as an early employee leading go-to-market; and Forter, the fraud prevention company, as employee number 10, leading go-to-market and business development during a period she describes as genuine product-market fit from day one. She later led Oracle’s startup ecosystem activity across Israel and additional regions, then led global transformation for a Mexican private equity firm. In 2020, she partnered with The Phoenix — Israel’s largest insurance company — to build and run a small early-stage fintech fund focused on baking real industry expertise into venture investing. She joined Viola FinTech after meeting her now-partners at a fund-of-funds event and, in her words, “dating” them for about a year before deciding to partner.

Viola FinTech and Viola Group

Viola Group operates four main investment strategies — three equity, one debt. Two funds focus specifically on Israel-related companies (one early-stage, one growth), investing as generalists with fintech and insurtech as part of a broader thesis. Viola Credit is a global debt fund. Viola FinTech is the only fund in the group that is global by definition — it invests some capital in Israel, but its primary focus today is the US market, having previously spent more time in Europe. Viola FinTech’s sweet spot is early stage: seed extensions and A rounds, with a first check typically around $5 million.

The Four-Layer Framework: Where Does Insurance Value Actually Live?

This is the intellectual centerpiece of the conversation, and it deserves careful attention because it reframes a debate most of the industry has been having in less precise terms.

Noam breaks insurance value into four layers, stacked from foundational to most defensible:

Layer 1: Capital

The most basic, essential layer — and, Noam argues, not where new entrants meaningfully compete. The barrier to entry is simply too high. This layer is largely off the table for startups trying to build a defensible position.

Layer 2: Underwriting

Historically, this is where insurtech startups went to create value — building better risk models through technology. Noam’s blunt assessment: AI is now commoditizing this layer rapidly, with important caveats.

Layer 3: Distribution

Sitting above underwriting, distribution is where getting the right product in front of the right customer at the right moment creates real, sustained value.

Layer 4: Trust

The top layer, and — in Noam’s view — the most defensible and least understood. Trust shows up in two distinct moments: at the point of purchase (am I getting the right offer, is this a fair deal) and at the point of claim (can I actually trust this brand to support me when something has gone wrong). Trust cannot be rented the way capital can, or hired the way underwriting expertise can. It has to be built, slowly, and it is exceptionally easy to destroy.

Why This Matters for How Insurance Companies Spend Money

Noam’s framework explains something that looks irrational on the surface: why a company like Geico spent close to $2 billion on marketing roughly a decade ago. It is not simply brand awareness. It is the ongoing cost of maintaining trust at scale — because trust, once damaged by a single bad experience, does not recover just because the brand spent heavily to build it in the first place.

The Controversial Take: Loss Ratio Obsession Is Necessary But Not Sufficient

Noam offered a genuinely provocative critique of how insurtech investing typically works.

Loss ratio is the number every investor fixates on — and for good reason, since it is easy to measure, empirically grounded, and directly tied to underwriting quality. But Noam’s argument is that this obsession creates a blind spot: distribution advantages and trust don’t get “underwritten” with anything like the same rigor, precisely because they resist easy quantification. There is no clean number for “does this company have a distribution moat competitors can’t replicate” or “how much trust has this brand actually built with policyholders.” Those questions matter as much as loss ratio — arguably more, over a long enough time horizon — but they don’t show up on the same spreadsheet.

AI and the Commoditization of Underwriting

Noam used cyber insurance as the clearest illustration of the pattern she sees playing out across the industry.

The Cyber Insurance Arc

Cyber insurance emerged because incumbent carriers simply lacked the data to underwrite the risk. Insurtech companies stepped in, built new underwriting capability through technology, and captured real value in that gap. Now, years later, incumbents have caught up on data, capacity is being pulled back in some cases, and the underwriting moat that once existed has largely closed. It is no longer a “new” risk requiring novel technology to price — it has become commoditized.

At Bay’s Sale to Munich Re

Noam and I discussed At Bay’s acquisition by Munich Re as a live example of this consolidation pattern. At Bay started as an MGA, later acquired a carrier, and had previously acquired and shut down another company called Relay. Its last funding round was a Series D in 2021 — roughly $200 million, with an additional follow-on bringing total raised closer to $300 million, at a post-money valuation around $1.3 billion. Five years later, the company sold at roughly half that valuation. Noam’s read: “all things considered,” a reasonably good outcome given how few 2021-vintage deals produced happy full cap tables — though clearly not the outcome the later-stage investors were hoping for when they wrote their checks at peak valuation.

The New Frontier: AI Liability as “The New Cyber Insurance”

Where Noam sees genuine underwriting moat potential today is in emerging, genuinely new risk categories — specifically AI liability. It has the same structural characteristics that made cyber insurance defensible in its early years: a risk incumbents don’t yet have the data or tools to price well, requiring both technology and specialized expertise to underwrite properly. That combination — new risk, real technical barrier — is precisely what makes a category defensible, at least until incumbents catch up.

Contextual FinTech: Why INSHUR Works and Not Every Embedded Moment Does

Noam’s “Contextual FinTech” thesis — first published as an article mapping the space — takes the broader concept of embedded finance and adds a missing ingredient: genuine user intent and timing.

The Distinction She Draws

Embedded finance, in its most basic form, focuses mostly on distribution and low customer acquisition cost, without much regard for whether the customer actually wants the product at that moment. Noam’s example of the failure mode: offering banking services inside a grocery retailer, where a shopper with kids in the cart has no meaningful intent to think about financial products at that specific moment.

Contrast that with INSHUR, a Viola FinTech portfolio company providing auto insurance for rideshare drivers, delivered directly inside the Uber driver onboarding flow. That is the exact moment a driver needs to make an insurance decision — genuinely contextual, genuinely timely, and a strong match between distribution strategy and customer experience.

Not a One-Size-Fits-All Solution

Noam was clear that this pattern does not generalize evenly across insurance products. Complex decisions — a mortgage, a home insurance policy — carry more inherent complexity than something like point-of-onboarding auto insurance for a single driver. When she originally mapped the contextual fintech landscape, insurtech was the vertical showing the strongest, most numerous examples of the concept actually working in practice.

AI Made Her Job Harder, Not Easier

This was one of the sharpest and most quotable moments in the conversation.

On the day-to-day operational level, AI tools genuinely help — less time spent on administrative and analytical grunt work. But Noam’s deeper point is about what has become harder to evaluate: the human judgment underneath every investment decision.

Every investor claims “team” is the most important factor. Noam’s refinement: what specifically matters about a team has shifted. Leadership capability, resilience, flexibility, willingness to keep an open mind, and the ability to adapt and move fast under changing conditions — these are now the qualities she is trying to assess before writing a check, and they are exceptionally difficult to evaluate in a short due diligence window.

She has watched some portfolio companies adapt rapidly to the AI-driven shift in their markets, and others struggle for far longer to find their footing. Her conclusion: “This is one question you can’t answer with Claude.” Human judgment, intuition built from operational experience, and pattern recognition from having managed teams — none of that is something an LLM can substitute for, no matter how good the tools get at everything else.

The Founder-Investor Relationship Is a Marriage, and 2021 Forgot That

Noam drew a direct parallel between how long it took her to “date” her eventual Viola FinTech partners — about a year — before formally joining, and what a fund relationship with a founder actually requires: eight to twelve years of partnership, often harder to exit than an actual marriage.

Contrast that with 2021’s investing environment: term sheets issued within 24 to 48 hours, deal cycles compressed to the point that investors genuinely could not get to know founders or fully understand the market opportunity before committing capital. Noam’s characterization: it was “almost like a lottery.” That compression, in her view, is a meaningful part of why so many 2021-vintage deals produced disappointing outcomes.

The Credit vs. Equity Distinction

Noam also drew a useful distinction between what creditors and equity investors actually want. Creditors want cash flow and collateral — they do not participate in the upside, so they optimize entirely for downside protection. Equity investors, by contrast, are underwriting vision and long-term potential — and when it works, the return reflects that risk. Early At Bay investors, by Noam’s estimate, saw roughly 20x returns on their initial investment, a very different outcome than the later-stage investors who came in near the 2021 peak.

Portfolio Spotlight: Advance

Noam highlighted Advance, one of Viola FinTech’s more recent investments, as a clean example of the fund’s broader thesis: infrastructure that captures real defensibility without ever touching underwriting risk directly.

Advance, founded by Omer and Gal, operates in the insurance space without underwriting a single policy. Instead, it provides the financial operations rails — the payment and money-movement infrastructure — that MGAs, brokers, and wholesalers need to actually run their businesses. Noam describes it as sitting at the genuine intersection of fintech and insurtech: real defensibility, without capital risk on the balance sheet.

What’s Next for Viola FinTech

Noam described a fund trying to embody the same flexibility and adaptability it looks for in the founders it backs. The broader market, in her framing, is mid-revolution, with many stakeholders still waiting to see how the current AI-driven disruption ultimately settles.

Viola FinTech continues actively scouting insurtech and fintech opportunities, while exploring newer business models — including, notably, the concept of AI-powered rollups specifically within the insurance industry. Nothing has been executed yet, but it is an active area of exploration. The fund’s broader thesis also spans fintech and healthcare, along with defense fintech — largely compliance-related technology. Viola FinTech re-evaluates and expands its thesis on a quarterly basis rather than treating it as fixed.

The Closing Answer: The Intersection of Distribution and Trust

Asked what the industry should be talking about more, Noam circled directly back to her four-layer framework: the intersection of distribution and trust is where AI is creating the most disruption and, correspondingly, the most new opportunity. Some companies are already talking about it. Her sense is there remains significant unclaimed territory for founders willing to build there.

Key Takeaways

• Insurance value sits in four layers — capital, underwriting, distribution, trust — and AI is actively commoditizing the middle two while making the top layer more valuable and more defensible
• Loss ratio obsession in insurtech investing is necessary but insufficient; distribution moats and trust are real value drivers that resist easy quantification
• Cyber insurance’s underwriting moat has largely closed as incumbents caught up on data; AI liability is emerging as the next genuinely new, genuinely defensible underwriting frontier
• Contextual FinTech succeeds when distribution timing matches genuine user intent — INSHUR inside Uber’s driver onboarding flow works precisely because the insurance decision and the moment of need are identical
• AI has made evaluating founders harder, not easier — leadership, resilience, and adaptability are now the primary signal, and they remain fundamentally unautomatable
• The founder-investor relationship functions like a long-term marriage; 2021’s 24-hour term sheet culture compressed that process in ways that produced avoidable bad outcomes