Why We Founded PolicyReview

There's a moment that sticks with me. I was talking to a bakery owner named Maria—third-generation family business, 12 employees, the kind of place where regulars don't need to order because the staff already knows.

She'd just had a claim denied. Water damage from a burst pipe. Should have been straightforward. Except her policy had an exclusion for "water that backs up or overflows from a sewer, drain, or sump pump." The pipe burst had caused a drain backup. Technically excluded.

Maria had been paying $8,400 a year in premiums. She'd never read the policy. Her broker had never flagged that exclusion. And now she was looking at $34,000 in repairs that nobody was going to cover.

"I thought that's what insurance was for," she said.

She wasn't wrong to think that. She was just wrong about what she'd actually purchased.

The Problem That Kept Showing Up

I've spent my career at the intersection of fintech, insurance, and marketing. I've seen how financial products get built, how they get sold, and how customers actually experience them. And commercial insurance kept showing up as a broken experience.

The pattern is always the same:

Business owner buys insurance. Business owner assumes they're covered. Business owner files claim. Claim gets denied or reduced because of something buried on page 73. Business owner says, "But I thought..."

It's not that insurance companies are predatory (though some are). It's that insurance policies are genuinely difficult documents to understand. A typical commercial policy runs 50 to 200 pages. It's written in legal language designed for precision, not comprehension. It contains dozens of exclusions, endorsements, and cross-references that modify each other in ways that aren't obvious unless you've spent years learning to read them.

The people who understand these documents—underwriters, claims adjusters, coverage attorneys—work for insurance companies. The people who need to understand them—business owners—are busy running businesses.

That asymmetry is where claims get denied.

The Math That Bothered Us

Professional policy review isn't cheap. We've seen quotes north of $400 per policy. A thorough review takes hours of focused attention from someone who knows what they're looking for—and that expertise commands real money.

At that rate, a small business with five policies could easily pay $2,000 or more for a complete portfolio review. That's real money for a small business. It's the kind of expense that gets pushed to "next quarter" indefinitely.

So most businesses don't pay it. They trust their broker, sign the paperwork, and hope for the best.

Hope isn't a strategy, but it's the one most businesses use.

I kept asking: why does understanding your own insurance have to be this expensive and complicated? You're already paying premiums. Shouldn't you at least know what you bought?

Finding the Right Partner

I knew this problem needed a technical solution. Not a better spreadsheet or a cheaper consultant—a fundamentally different approach to how policy review gets done.

That's when I connected with Aaron.

Aaron's background is tech and startups. He's built products, scaled systems, and has the engineering instincts to know what's actually possible versus what's just a pitch deck fantasy. When I explained the policy review problem to him, his first question wasn't "is this a good market?" It was "why hasn't anyone solved this yet?"

The answer, historically, was that reading dense documents and extracting meaning from them required human expertise and human attention. Those are expensive and don't scale.

Aaron's response: "That answer is about two years out of date."

What Changed

Modern language models can read. Not in a gimmicky, keyword-matching way—they actually comprehend text. They can work through a 150-page policy, identify the coverage grants and exclusions, find the endorsements, and explain what they mean in plain language.

We know because we tested it. Extensively.

We fed policies to AI systems. We compared their outputs to professional human review. We looked for the gaps—the things the technology missed that a trained expert would catch.

The gaps exist. AI isn't perfect. But the gaps are smaller than we expected, and they're getting smaller every month.

More importantly, the technology is fast and cheap. A review that takes a human consultant several hours takes AI about three minutes. The cost drops from $400+ to almost nothing at the margin.

That changes the math completely.

What We're Building

PolicyReview is a tool that reads your commercial insurance policy and tells you what's actually in it.

Not the declarations page summary. Not the broker's pitch. The actual policy—every coverage, every exclusion, every endorsement, every limit and sublimit and condition.

We extract the information. We organize it so it makes sense. We flag the gaps and issues. We explain what we find in language that doesn't require a law degree to understand.

The goal is simple: you should know what you're buying before you need to use it.

What We're Not Building

We want to be clear about what PolicyReview isn't.

We're not a replacement for insurance brokers. Brokers do things we don't—they shop the market, they negotiate with carriers, they advocate for you when claims go sideways. Those are valuable services.

We're not giving insurance advice. We don't tell you what coverage to buy or whether your limits are adequate for your specific situation. That requires understanding your business in ways that a document review can't provide.

We're not guaranteeing perfection. AI can make mistakes. Policy language can be ambiguous. Complex coverage questions sometimes require human judgment that no technology can replicate.

What we are is a tool that helps you understand what you've purchased. Think of it as a translation layer between insurance-speak and human-speak. A sidekick that gets you 90% of the way there, even if commercial insurance isn't your area of expertise.

The Bet We're Making

We're betting that most businesses would make better insurance decisions if they actually understood their coverage.

We're betting that the friction isn't willingness—it's access. Business owners aren't ignoring their policies because they don't care. They're ignoring them because understanding requires expertise and time they don't have, and professional review costs more than feels justified until something goes wrong.

We're betting that technology can close that gap. Not perfectly, but well enough to matter.

And we're betting that there's a market for "good enough to matter." Not just enterprise risk management departments with six-figure budgets. Regular businesses. The bakery owner. The electrical contractor. The marketing agency with 15 employees.

The businesses that are currently hoping their insurance works because they've never had an affordable way to actually check.

Why Now

Timing matters for startups, and the timing here is specific.

Two years ago, the AI technology wasn't good enough. Language models could extract keywords and match patterns, but they couldn't really comprehend complex conditional language—the kind that makes insurance policies what they are.

Two years from now, this space will be crowded. We're already seeing competitors emerge. The opportunity is obvious to anyone paying attention.

Right now, the technology works and the market is nascent. That's a window, and windows close.

We'd rather be early and have to educate the market than late and have to differentiate in a crowded field.

What Success Looks Like

If we get this right, fewer conversations end with "I thought I was covered."

Fewer claims get denied because of exclusions nobody knew about. Fewer businesses discover gaps at the worst possible moment. Fewer people pay premiums for coverage that doesn't match their actual risks.

That's the outcome we're optimizing for.

The business model matters—we need to be sustainable to have impact—but the business model follows from the outcome. If we genuinely help businesses understand their coverage, the revenue takes care of itself.

The Invitation

If you run a business and you've never really read your insurance policy, you're not unusual. You're normal.

But normal isn't the same as safe.

We built PolicyReview because we believe you deserve to know what you're buying. Not because we want to scare you into buying more coverage—that's not our business model. Because informed decisions are better decisions, and right now most insurance decisions aren't informed.

Give us your policy. We'll tell you what's in it.

That's the pitch. No games, no upsells, no strings. Just clarity about something that matters more than most businesses realize.

Don Halliwell brings years of experience across fintech, insurance, and marketing—he's seen how financial products get built, sold, and experienced by real customers. Aaron Reese is a technical co-founder with deep roots in tech and startups, focused on building products that actually work at scale. They founded PolicyReview because understanding your own insurance shouldn't cost $400 and a law degree.