YAS MicroInsurance's Founder: "I'm Not Insuring the Person, I'm Insuring Their Data"

Andy Ann, CEO and co-founder of Hong Kong-based YAS MicroInsurance, describes his company's core innovation in a single reframe: instead of insuring a person, YAS insures a unit of activity -- kilometers run, rides taken, minutes on a mountain trail. "Our model is frequency," he said. "The model is kilometers. The model is time." That distinction sounds semantic until you follow its consequence to the pitch YAS makes reinsurers: a city's population is a fixed, bounded number, but the volume of kilometers its residents might run in a year is not. "The pie extends," as Ann put it -- the addressable market scales with behavior rather than headcount, which is precisely the pitch that gets a reinsurer past their default reaction to a $2-3 premium: "don't talk to me unless it's a million dollars."

Insurance sold by the kilometer

The clearest expression of that thesis is YAS's running insurance product, which sells coverage in literal kilometer blocks -- a customer buys, say, 500 kilometers of cover, and it depletes as their tracked runs accumulate. Someone who doesn't run at all generates no exposure and pays nothing further; someone training heavily consumes their block faster and tops up. The same logic extends to event-specific coverage: a runner who signs up for a marathon can be insured not just on race day, but from the moment they register, covering months of training beforehand -- with YAS even able to refund the event's entry fee if the runner doesn't show up on the day. "It's not only a day of event protection," Ann said. "It's actually by quarter, by week, by month, by kilometers."

Building products from unusually specific customer research

YAS's product design process leans heavily on direct, granular customer interviews rather than top-down actuarial assumption. For its road biking product, the company recruited an early customer pool -- essentially a Kickstarter-style founding cohort -- and interviewed them specifically about what they cared about protecting. The answers split sharply by gender: men, who'd typically spent upward of $2,000 on their bikes, prioritised protecting the bike itself; a clear majority of women cited facial injury from a fall as their top concern. That research directly shaped a three-part product covering the rider's injuries, their personal belongings (phones, AirPods, gear), and the bike itself -- a design a generic, single-dimension policy would have missed entirely.

Pricing went through the same direct-research process. Rather than setting a premium and testing demand indirectly, the team asked prospective customers a willingness-to-pay ladder directly: would you pay $6 per ride? (Never.) $4? (Maybe.) $3? (Yes, definitely -- "it's just a bottle of water, a cup of coffee.") That $3 anchor point, framed against a familiar small daily purchase rather than an abstract insurance premium, became the product's actual price.

The research also surfaced sharply different psychological entry points across customer segments. Hikers, Ann found, tend to buy coverage for their families rather than themselves -- they see themselves as the protector, not the protected. Ride-hailing passengers, by contrast, typically don't think about insurance at all until a specific trigger moment: his example was a late-night ride where a driver visibly speeding after the passenger has had a couple of drinks suddenly makes the risk concrete and immediate. Each persona, in his account, needs an entirely different message and moment to actually convert.

A distribution model built on embedding, not selling

Because YAS's product depends on frequency and data rather than a one-time underwriting decision, its distribution strategy looks structurally different from a conventional insurer's. Coverage is embedded directly into wearable devices, into e-commerce transactions (a pair of running shoes sold with three months of running-injury cover attached, renewable after that), and into a transit partnership with a bus operator carrying roughly 8 million daily trips -- putting insurance directly inside high-frequency transactions rather than asking customers to separately seek out and purchase a policy.

Alongside embedded distribution, YAS invests deliberately in building category-specific communities: partnerships with major races and events (Spartan races, mountain hiking events, marathons), and a full ecosystem of bike shops, repair shops, apparel brands, key opinion leaders, and even government-affiliated cycling associations around each product line. Ann was candid that this is genuinely hard, unglamorous work -- "forming a community... you have to generate content, you have to add value, you have to conduct events" -- but argued that difficulty is precisely the point. A generic app is trivial for a competitor to copy; a genuinely engaged, purpose-built community around a specific activity is not. "Your entry barrier is so high that no one can come in," he said, framing community-building as a durable moat rather than a marketing nice-to-have.

Why NFT insurance is harder than it looks

One of YAS's more ambitious bets is Nifty, an insurance product for NFTs and digital assets -- built on the observation that under 1% of the value circulating in the crypto and Web3 economy is insured against loss, theft, or smart-contract exploits. YAS ran an initial pilot with a roughly $2 million risk pool, working with artists and foundations across London, Paris, the US, and Hong Kong.

The first version, though, exposed a structural design flaw worth understanding for anyone building insurance products around volatile digital assets: premiums and claims were converted between crypto and fiat currency rather than held in a crypto-denominated pool, which created a mismatch problem given how sharply an NFT's market price can move between when a policy is issued and when a claim is actually filed. YAS's version 2 redesign specifically separates two distinct risks -- exploits occurring during the earliest transaction blocks after a transfer, and loss or theft of an asset already held in a wallet -- rather than treating digital asset risk as a single, undifferentiated category. Notably, Ann cited the UAE's comparatively flexible and forward-looking crypto regulatory environment as a specific reason YAS is exploring expansion there, alongside Malaysia and Vietnam. He was also explicit about rejecting fully decentralized, unregulated peer-to-peer insurance models on principle: "Something has to come into regulate... [to ensure] there is a captive pool that [is] able to pay off" -- a deliberate choice to move slower through proper licensing rather than launch faster without it.

Recession as recruiting opportunity, not just a threat

On the funding environment -- this conversation took place amid a sharp 2022 rate-hiking cycle, with venture firms publicly warning portfolio companies to brace for a rougher period -- Ann's framing was notably contrarian. Rather than treating the downturn purely as a threat, he argued a difficult market is often the best time to build: talent laid off from larger, less disciplined competitors becomes available, product categories abandoned by companies that couldn't sustain them become open again, and the market stops rewarding "hype stories" and pure valuation momentum in favour of genuine revenue and product-market fit -- criteria a disciplined, data-driven company like YAS was already built around.

On the metrics side, Ann tracks CAC and lifetime value closely, but flagged a category-specific nuance worth noting: while traditional term life insurance carries a 15-20 year LTV horizon, YAS's own micro-duration products have a structurally shorter 3-5 year horizon -- a fundamentally different retention and monetisation math that any usage-based insurtech needs to plan around rather than borrowing life-insurance-style assumptions wholesale. He also described active fraud-detection work built directly from claims data -- flagging a specific case where a single customer filed suspicious claims repeatedly despite police reports attached, prompting the company to build out dedicated internal investigation tooling.

What this means for the region

YAS's core thesis -- that reframing insurance around frequency and behavioral data rather than static population segments unlocks both product design and reinsurer appetite that a conventional model can't reach -- is directly applicable to GCC insurtechs targeting similarly bounded, small national markets. A UAE-based usage-based insurer facing the same "our population is too small" objection from underwriters has, in YAS's kilometer-based running product, a concrete template for reframing the pitch around activity volume rather than headcount. Equally transferable is the underlying research discipline: granular, segment-specific customer interviews -- down to gender-specific risk concerns and direct willingness-to-pay testing -- produced a materially better product than any top-down actuarial assumption would have, a method any regional insurtech building for an equally under-researched Gen Z and millennial customer base could replicate directly.

This post draws on the FS Brew episode 23: The TAM for micro-insurance is huge- Interview with Andy Ann.