📕 Startup term you should know

Ever heard of Product-Market Fit Score?

Want to know if people love your product? Send them a survey and ask them a simple question: "How disappointed would you be if this thing just... disappeared?" If more than 40% say "very disappointed," congrats, that’s one more signal that you’ve got strong product-market fit.

My insider scoop: Sean Ellis developed the "40% test" after analyzing hundreds of startups. For angel investors, this metric is more reliable than vanity metrics like user growth. However, Rahul Vohra warns that "PMF surveys only work if you survey actual users, not just people who signed up." Marc Andreessen notes that "product-market fit is binary - you either have it or you don't, and you'll know when you have it."

📰 Today's topic: The “boring” startup has a better moat than the cool one

Let's run a quick test. You've got $5,000 and two companies in front of you.

Company A is an AI tool that spins up gorgeous presentations in about ten seconds. The demo made you smile. You understood the pitch in one sentence. Company B does something with compliance plumbing inside digital health, and (be honest) you started checking your phone halfway through the explanation.

Almost every new angel reaches for Company A. We'd nudge you toward B.

Think about what the big AI labs did in the first quarter this year. One of them shipped something like 120 features in 90 days. Ninety days! And it puts a hard question in front of any early-stage bet you make: what stops a giant model from doing what your company does, for free, next quarter?

The answer comes down to one word. Is the company horizontal or vertical?

Horizontal companies cut across everything. The slide tool, the writing helper, the do-it-all assistant. They're easy to grasp and easy to love, and that's the trap. The same broad usefulness that makes them appealing makes them easy to copy. When the next model drops a hundred new features, a few of those features might just be your portfolio company's entire reason for existing.

Vertical, regulated companies are harder to flatten. Fintech. Digital health. You can't roll out of bed and code your way into those worlds. You need licenses. You need someone who actually understands regulatory workflows that took years to learn. AI can write code in a blink, but it can't grant itself a banking charter or fast-forward through the part where a human has to know the rules cold. That friction is the moat.

We've felt this firsthand. We hold a position that’s up 103x, and it's one we watch most nervously - precisely because it's horizontal. Real money, great company, but more exposed when the ground keeps moving. Our “duller” regulated bets are the ones that just keep chugging.

The unsexy verticals everyone skips usually come with friendlier valuations, because there's no crowd of angels elbowing in. So you get a better price on a sturdier business.

– Brian from Angel Squad

🍫 A snack for the road: Automating the wrong things

Every company's using AI to hire faster. Resume screens, scheduling, the rejection email...all of it is automated now. It got faster and cheaper.

But bad hires kept coming at the exact same rate.

Half of job seekers got rejected last year without a single word from a human.

So the sorting got faster but the hiring didn't get better.

That's the premise of a virtual event next week that I'm hosting called "Automating the Wrong Things."

Point AI in the wrong direction and it just delivers you to the wrong hire faster.

Mappa is making the case that 30 seconds of someone's voice reveals what no resume, interview, or gut read ever could.

Not faster hiring → clearer hiring. Fewer people pushed through a funnel, but much better data about each one.

🎥 I'm hosting this Thurs, 6/18 at 10AM PT.

Overheard in SF…probably

“We're pivoting from dog walking to quantum computing. The skills are totally transferable, both involve moving things from point A to point B.”