80% of the People Who Bought His First Policy Were Otherwise Completely Uninsurable

Jonathan Gonzalez’s path into insurance runs through the worst kind of personal crisis. In 2017, having just moved back to Puerto Rico to care for his mother — who uses a wheelchair and was medically fragile — Hurricane Maria hit. It took him three days just to reach her, deep inside the island. The wheelchair ramp had been blown out, there was a hole in the house, and the very first thing he did was ask whether she had insurance and how to file a claim. It took over a year to get an adjuster to even show up — and the family couldn’t repair the ramp or the hole in the meantime, because the adjuster needed to inspect the damage first. After the adjuster finally came, it took another six months of back-and-forth with the insurer. The outcome: zero payout.

His first question wasn’t about insurance at all — it was whether this was simply bad luck, a uniquely bad policy, or something structurally common. Asking around, he quickly learned it was a near-universal story, both in Puerto Rico and, once he and his future co-founders looked outward, in disaster-affected regions generally — the media covers the disaster itself extensively, but almost never the aftermath of people still fighting for a payout months or years later.

In Episode 84 of InsurTechTalk, Jonathan and I covered how that personal disaster became a parametric insurance company operating across the Caribbean and Latin America, why his first product uncovered a massive coverage gap nobody had been looking for, and his precise, and genuinely useful, definition of what “embedded insurance” actually means.

About Jonathan Gonzalez

Jonathan Gonzalez is CEO and co-founder of Raincoat, a parametric insurance infrastructure company based in Puerto Rico, building embeddable climate risk products and the technology stack that powers them for insurers, reinsurers, MGAs, and governments across Latin America and the Caribbean. Before Raincoat, he founded an IoT startup in New York. Raincoat’s team is deliberately technology-heavy — data scientists and engineers from Microsoft, Google, Oracle, and NASA — paired with genuine insurance expertise, including Chief Insurance Officer Luis, a five-time insurance company CEO. Raincoat raised a $4.5 million seed round from investors including Anthemis and SoftBank, alongside local Puerto Rico investors.

Why Parametric, and Why It Wasn’t Already Mainstream

Coming from a software background, Jonathan’s engineering instinct led him directly to parametric insurance once he started researching alternatives to traditional claims-based coverage — the concept of breaking insurance down into measurable parameters felt immediately native to how software engineers already think. His next question was the obvious one: if parametric insurance has existed for two decades, why hadn’t he — or his mother — ever encountered a policy?

Meeting with insurers, reinsurers, and everyone across that value chain, he found the same recurring set of blockers:

• Risk and detection model gaps — the underlying data and modeling infrastructure many carriers simply didn’t have in place
• Infrastructure complexity — a genuinely simple parametric product doesn’t need much, but anything involving image recognition or broader machine learning quickly runs into real big-data infrastructure requirements most insurers weren’t equipped for
• Legacy policy administration systems — even carriers with good risk models and adequate data were often structurally unable to automate claims payment, because their core systems simply weren’t built for it

His articulation of the ideal outcome crystallizes exactly what Raincoat set out to build: his mother receiving a text message within days of the hurricane confirming a fixed payout, deposited directly to her bank account — no adjuster, no year-long wait, no manual back-and-forth.

Building the Infrastructure Layer, Not the Distribution Layer

Raincoat’s positioning is explicitly infrastructure-first: rather than trying to sell insurance directly, the company builds parametric products and the underlying technology stack, then delivers that packaged product to whatever channel — an insurer, reinsurer, bank, or government — wants to distribute it. Jonathan drew a direct parallel to affinity-based distribution models discussed elsewhere in the podcast: Raincoat’s mission (protecting people from climate disruption at scale) requires reaching every possible channel, which means building genuinely embeddable products rather than trying to own the customer relationship itself.

Why the Company Had No Website — On Purpose (Sort Of)

At the time of this recording, raincoat.com was essentially bare — a scenario Jonathan was candid wasn’t a deliberate stealth strategy so much as an accident of how organically the company grew. Raincoat never set out with startup ambitions from day one; it began as an urgent, personal problem to solve, and formalized into a company only afterward. Its growth to that point had been almost entirely word-of-mouth — one successful regional deployment leading directly to an introduction in the next region, repeatedly, without ever needing active marketing. The recent seed round, he noted, marked the inflection point where that changed — this conversation was among the first public discussions of the raise itself.

”These People Have to Be Evil” — And Why That Turned Out to Be Wrong

Jonathan was disarmingly honest about his initial emotional read on the industry after his own family’s experience: his gut reaction was that the people who denied his mother’s claim had to be acting in bad faith. What he discovered instead, working closely with Luis (his eventual Chief Insurance Officer, who had been running an insurance carrier in Puerto Rico as CEO during the very same hurricane), was more structurally troubling: the failure wasn’t any single bad actor, but a systemic misalignment of incentives across the full stack — the person structuring the risk thinking about one set of priorities, the person drafting the legal contract thinking about another, the salesperson focused purely on closing the sale — with nobody holding end-to-end visibility or accountability for the actual customer outcome.

Luis’s own account of living through the same disaster from the carrier side was, in Jonathan’s telling, genuinely revealing: moments where Luis personally wanted to simply issue a check on the spot, blocked entirely by regulatory process. The volume mismatch compounded the problem structurally — a carrier that might normally employ 10 claims adjusters suddenly facing 100,000 claims simultaneously has no realistic way to scale that capacity fast enough, regardless of good intentions.

The First Product: A Micro-Insurance Policy That Revealed a Hidden Gap

Raincoat’s first-ever product was a standalone parametric micro-insurance policy in Puerto Rico — notably the first fully admitted micro-insurance policy filed with the NAIC in that category. It wasn’t designed to replace or bundle with a traditional property policy; it paid a fixed cash amount based purely on measured wind speed intensity, calibrated against the kind of excess expenses (food, gasoline, temporary needs) a household actually incurs after a serious hurricane.

The genuinely surprising finding once the product launched: roughly 80% of buyers purchased it not as a supplement to existing coverage, but because it was their only available safety net — they were otherwise completely uninsurable through any traditional product. This was despite Puerto Rico having relatively high insurance penetration by Latin American standards (Jonathan cited over 12%) — meaning a real, sizable protection gap existed entirely underneath what standard penetration statistics were capturing. Jonathan connected this directly to comments from the CEO of PeakRe at a recent Monte Carlo conference: the industry’s real need isn’t more sales pressure on existing products, but genuinely new product categories addressing previously unaddressed or newly emerging risk — an “expand the pie” framing rather than a zero-sum competition for existing premium.

MGA Status: It Depends Entirely on the Channel

Asked directly whether Raincoat operates as an MGA, Jonathan’s answer reflected the company’s deliberately flexible structural approach: in Puerto Rico, yes. In other markets, it depends entirely on what the specific channel requires. His example: in Mexico, Raincoat works with the Mexican government and a consortium of reinsurers (Swiss Re, AXA, Munich Re, with Guy Carpenter as reinsurance broker) through a government-sanctioned policy structure that doesn’t require an MGA layer at all. Jonathan’s summary of Raincoat’s actual role, regardless of the specific structure required: build the product, bring the capacity, build the infrastructure, integrate with the channel’s core systems, and deliver it fully packaged — what he calls “abstracting the chaos” that makes reaching underserved populations (his example: small-scale farmers) genuinely difficult for most players to attempt at all.

Product Footprint: Puerto Rico to Jamaica to Mexico to Colombia

Raincoat’s deployment sequence, as of this recording: the founding micro-insurance policy in Puerto Rico, followed by an agricultural product line with an insurer in Jamaica, an active deployment in Mexico, and a new deployment in Colombia going live shortly after this conversation, with several others in process. Jonathan’s broader observation on market effect: in every market Raincoat has entered, its product was the first of its kind in that specific market — and in Puerto Rico specifically, the original single micro-insurance policy category has since grown to roughly five distinct product types distributed through different channels, a wave he’s glad to have helped start without claiming sole credit for where it’s since gone.

Defining Micro-Insurance and Embedded Insurance Precisely

Two definitions Jonathan offered are worth capturing directly, since both terms get used loosely across the industry.

Micro-insurance, in his working definition: policies with genuinely low coverage limits — not competing with traditional insurance products in scope or intent, with the specific dollar threshold varying meaningfully by market (a few thousand dollars means something very different in different economies). Most Raincoat products end up being micro-insurance essentially by consequence of being embedded and mass-consumer-focused, not because micro-insurance was the explicit design target.

Embedded insurance, his sharpest and most quotable framing: insurance that isn’t marketed at all — coverage integrated as close as possible to the actual insurable interest, ideally as simple as a checkbox at the point of an unrelated transaction, or bundled by default. His illustrative example: e-commerce return policies function essentially like an insurance product already built into pricing that nobody thinks of as insurance — he floated the thought experiment of an alternate universe where an Amazon order explicitly offered “pay $1.53 for return protection” as a visible line item, versus the current reality where that risk is simply priced into every transaction invisibly. His view: insurance in general should work this way — coverage arriving automatically at the moment of the underlying transaction (buying a house, renting a car) rather than requiring a separate, disjointed shopping process — which is part of why the old industry adage “insurance is sold, not bought” persists as strongly as it does.

Culture and the “Hope Factor” in Purchasing Behavior

A genuinely distinct insight from Luis’s decades of Latin American market experience: religious and fatalistic cultural framing — a strong “everything is in God’s hands” mentality — measurably affects insurance purchasing behavior in parts of the region. People holding that worldview are statistically less likely to purchase coverage, not from lack of access or awareness, but from a genuine belief that outcomes are outside their control to influence through financial preparation. Jonathan noted this as a real, observed factor in market education and product design decisions, not an abstract cultural generalization.

Advice: The Untethered Soul

Asked for a closing recommendation, Jonathan pointed to The Untethered Soul: The Journey Beyond Yourself — a book he described as unexpectedly good despite sounding, in his own words, “a little wishy-washy,” with genuinely Buddhist framing he found broadly applicable regardless of industry or whether someone works in a startup at all.

Key Takeaways

• Raincoat’s founding insight came directly from a personal disaster — a year-long, ultimately unsuccessful claims process after Hurricane Maria — that revealed the failure wasn’t bad faith but systemic misalignment across the full insurance value chain
• Legacy policy administration infrastructure, not lack of risk modeling sophistication alone, is frequently the real blocker preventing carriers from offering genuinely fast, automated parametric payouts
• Raincoat’s first product uncovered that roughly 80% of buyers were otherwise completely uninsurable through traditional channels — a real protection gap hidden underneath seemingly healthy overall market penetration statistics
• New, genuinely novel insurance products addressing previously unaddressed risk expand the total addressable market rather than just competing for existing premium — a distinction worth applying broadly across climate and parametric product design
• MGA status and distribution structure should follow what each specific market and channel actually requires, not a fixed template applied uniformly across every deployment
• Embedded insurance, properly defined, means coverage that isn’t marketed at all — integrated as a checkbox or default bundle at the point of an unrelated transaction, structurally different from even a well-executed direct sales motion
• Cultural and religious framing measurably affects insurance purchasing behavior in ways that pure access or product design improvements alone can’t fully address — a genuine market education variable, not just a distribution problem