Ask for an opinion, not a sale: how to find beta users for a B2B product before product market fit

Sunny Ray · Jun 6, 2026 · 5 min read

Pre-product-market-fit cold email fails because founders ask for the wrong thing: a sale instead of an opinion.

To find beta users for a B2B product, email the people who'd use it every day, not the executive who'd sign for it. Ask for feedback in two lines, not a demo. Qualify them before they reach your calendar, so every call is with someone who fits. Then switch to selling only when the feedback starts repeating itself. Do it in that order and cold email becomes your fastest discovery channel instead of a waste of your week.

It doesn't attract the right people

On June 4 I had a follow-up call with a founder who had recently been a guest on my show. His company builds tools for AI engineers, and he's pre-product-market-fit by his own account, with a formal PMF assessment planned for the end of August.

I asked whether they'd tried outbound. His answer was honest:

Yeah, we have. It's just. It doesn't attract the right people.

Then he added the part that tells you the real cost. They'd had a whole sales team. They didn't anymore. It was too early for one.

His diagnosis of the cold email: it pulled in the wrong audience, people too early themselves to give useful feedback, and there was no qualification step, so the team spent hours on calls with people who were never going to be users. He'd concluded cold email doesn't work for his stage. I think he'd concluded the wrong thing.

Why pre-PMF cold email attracts the wrong people

Most pre-PMF outbound is a sales email wearing a disguise. It names the product, lists features, and asks for a call. That email attracts two kinds of people: the curious who reply to anything, and the polite who take the call and never use the product. Neither tells you anything true.

Earlier this year I saved a post on X that said the worst thing in sales is pitching the solution too early. Before product market fit, any pitch is too early. You don't have a solution yet. You have a hypothesis. And a hypothesis needs critics, not customers.

Target the user, not the buyer

The first fix is who you email. His ICP wasn't the CTO. It was the individual contributor: AI developers and engineers working under an AI lead, the people who'd open his tool every day.

That matters more pre-PMF than at any other stage. The CTO is the buyer later. Right now, the CTO can't tell you whether the workflow is right, because they don't live in it. The engineer can. The engineer will also tell you, bluntly, which part is annoying, which part they'd pay for out of their own pocket, and which part they'd never touch.

So build the list around the user's title, not the budget holder's. It's a smaller list, it's easier to reach, and every reply is worth more.

The two-line feedback ask

The second fix is the ask. What worked better for him than his cold email was the simplest thing on the call: a two-line message asking for product feedback, the way early-stage founders have done it for years. Something like:

We're building a tool for engineers who [specific workflow]. Would you look at it for ten minutes and tell us what's broken?

That's it. No features, no company history, no meeting link. The ask is small, it flatters the expert, and it filters for people who care about the problem, because nobody else bothers to answer.

Notice what's missing: any reason to say yes other than the work itself. That's the point. Pre-PMF, you want the people who reply because the problem bugs them.

Qualify before the calendar, not on the call

The third fix is the one that saves the most time. He told me direct qualification on LinkedIn worked better than his old email flow. Before anyone got a call, they'd confirm their role and what they were working on. That one step cut out the people who were too early or in the wrong job.

You can do the same with email. One qualifying question in the reply, or two or three on the booking page: what's your role, what are you building with, how often do you hit this problem. If someone won't answer three questions, they weren't going to give you thirty minutes of honest feedback anyway.

His team had wasted hours on calls with unqualified respondents. That's the expensive part of bad outbound, not the emails. Emails are cheap. Founder hours aren't.

When to switch from feedback to selling

He wasn't ready to buy outbound from us, and at his stage that's the right call. He's a few months early. The plan is to talk again in September, after his PMF assessment, which is the moment the job changes.

Here's how you know you're there. The feedback starts repeating. Users stop telling you what's broken and start asking when they can have it for their team. Someone asks about pricing before you mention it. That's when the ask flips from tell us what's broken to want to try it with your team, and that's when the buyer, the CTO you skipped, goes back on the list.

Run the same machine in both phases. Same lists, same infrastructure, same qualifying step. Only the ask changes.

A better feedback channel: interview them

One more option, because it's the one I'd use. Instead of asking engineers to review your product, invite them onto your show to talk about how they work. Users love talking about their own workflow, and a recorded interview is the deepest customer discovery you'll ever get. You'll hear the problem in their words, which is exactly the language your landing page needs.

If that sounds like your next quarter, interview your ICP on your own show.

Frequently asked questions