If Your Answer Is “Everyone,” Investors have Stopped Listening

“Our target customer could be anyone with a SaaS business.”

That was the answer when I asked a cybersecurity founder who his ideal customer was. He’d been in the industry most of his career.

And the annoying part is that he wasn’t wrong, the way a horoscope isn’t wrong when it tells you you’ve been thinking about something lately.

Almost every software company does need security. He could name dozens that could plausibly buy what he’d built, and he’d have been right about all of them.

Why does an answer that is factually reasonable make a company sound less investable?

Compare it to an answer I got from another founder in roughly the same market. He sold to SaaS companies moving upmarket. Specifically, the ones whose first genuinely large customer had just sent a security questionnaire they couldn’t complete. What had been a nice-to-have for three years became the thing standing between them and the biggest contract they’d ever signed.

He hasn’t identified a better market.

It’s the same market.

What he’s revealed is that he’s learned something about it: who buys, when they buy, and why the problem stops being optional. There’s a mechanism in his answer. There is none in the first.

Which points at something other than the conclusion everybody reaches for. The quality of an (Ideal Customer Profile) ICP isn’t determined by how narrow it is. It’s determined by how much information it contains.

That distinction matters because “niche down” is bad advice dressed as good advice. “European SaaS companies with 50-250 employees” is far narrower than “anyone with a SaaS business” and contains almost as little. It’s a filter. It tells you where to point a list. It doesn’t tell you why anyone signs. A narrow ICP with no mechanism inside it is just a smaller version of the same non-answer, and investors read it the same way.

So where does the information come from?

It comes from being told no. Precisely, repeatedly, and in ways that remove possibilities. You start with a large number of plausible truths: this buyer, this size, this channel, this price, this pitch. Then reality starts deleting them. Not this title. Not before this trigger. Not through this route. Not at this number. What’s left at the end isn’t a segment you chose. It’s what survived.

Which produces an inversion I’d hold onto: a good ICP is largely a record of everyone you’ve learned not to sell to. Look again at “anyone with a SaaS business” and you can see exactly what’s missing from it. No subtraction has happened. Nothing has been ruled out. It is the answer of someone who has never been told no in a way that counted.

And here’s where it gets uncomfortable for the founder who’s actually good at this, because he has been told no. Hundreds of times. He has heard more of this market say no to his face than any investor he’ll ever meet.

But ask him why the last five deals died and you get five reasons. That buyer was mid-restructure. That one was three years into a contract with the incumbent. That CISO was six weeks into the job and was never going to champion anything. Every one of those is true, and because it’s true, the deal gets closed out and nothing gets written down.

Call it the pre-explained no. Experience hands you a credible reason for every individual rejection, so no rejection ever has to sit next to another one.

Put those five explanations on the same page, though, and you might find all five were the same size of company. Or that none of them had a deadline. Or that in all five the person who first got excited was never the person who signs. That’s the ICP. It was in the losses the whole time; it just never had to be in one place at once.

That’s what the investor is reaching for when they ask who wants this. They’re not checking whether you can produce a respectable customer persona; they can build one of those from a research report on the flight home.

Anyone can look at cybersecurity and work out that software companies need security. If you’ve spent eighteen months inside the problem and your answer is still essentially that, then the honest reading is that no information has been acquired. The ICP question is cheap to ask and hard to fake, which is why it comes early. What they’re listening for is the residue of contact with reality.

And they’re not asking out of intellectual curiosity. They’re asking because the answer to that question is the beginning of a sales motion.

If you’ve learned that companies buy when a specific thing happens to them, you know when to approach. If you’ve learned that one title consistently champions this internally while another takes the meeting, agrees with everything and disappears, you know whom to approach. If you’ve learned that one kind of company closes in six weeks and another takes nine months, you know where the next salesperson should spend their time, and roughly what to expect from them by the end of the quarter.

The same information that makes your answer convincing to an investor is what makes your GTM less expensive to run. The investor isn’t applying an arbitrary fundraising test. They’re checking for something the business needs whether or not anyone ever funds it.

Which is what makes “anyone with a SaaS business” such an expensive answer. It doesn’t only make the investor’s job harder. It makes yours harder, every week, in ways that don’t announce themselves as an ICP problem: the outbound that gets replies but no meetings, the pipeline that looks fine until the third month, the good salesperson who somehow doesn’t work out.

One last thing, because it’s usually the real objection. Founders resist a specific ICP because it feels like agreeing to a smaller market, like signing away the companies outside the definition before you’ve even tried them.

But a good ICP doesn’t make your market smaller. It makes your sales motion less random. You can still sell to the company outside it, happily, when they come to you. You just stop building your entire go-to-market around the hope that everyone behaves the same way.

Investor Translation

Founder says: “Anyone with this problem.” Investor hears: You haven’t been told no precisely enough yet.

Founder says: “We’re seeing interest across several verticals.” Investor hears: You can tell me where there’s interest. You can’t yet tell me where it converts.

Founder says: “We don’t want to limit ourselves too early.” Investor hears: You’re protecting optionality instead of following the evidence.

One Question

What do your last five lost deals have in common?

Not why you lost each one. What they have in common.

If the pattern isn’t obvious, send me the five reasons. I’m curious what shows up when they’re put next to each other.

Worth Keeping

"All models are wrong, but some are useful.", George Box

An ICP is a model. It will be wrong about somebody: a company you ruled out will turn up one day and buy anyway.

That’s the cost. It’s cheaper than a model that includes everyone and therefore predicts nothing.

Clipped

I was going through the YC library recently. Great way to spend an afternoon on holiday, btw.

Video link: Successful founders are OK with rejection

YC Partners discuss the customer who loves your product, can’t buy this quarter, but promises they’ll be in touch.

Is that a maybe? Or is it just a well-disguised no?

Worth ten minutes for anyone trying to get better at customer conversations, and at learning from rejection.