When your pipeline stalls, check which conversation you're having

This tier 1 bank-integrated fintech had been around for over 10 years. They had a small sales team, they were talking to customers — closing calls, account check-ins, support tickets, renewal conversations, doing all the things. By most standards, they were doing what you're supposed to do in running a business. But their pipeline had stalled, sales were down 23% over 18 months.

So we ran a Bullseye Customer Sprint. We built a hypothesis, developed a one-liner, and started interviewing SaaS CFOs who needed real-time visibility into processing fees with modern dashboards. By the third conversation, we heard a CFO say, "I don't care about saving 50 basis points if my team has to spend 20 hours a month manually fixing broken data in NetSuite. That's not a savings problem. That's an audit problem. I kick it to my controller."

Two more CFOs said versions of the same thing. We went back through earlier interviews and found a fourth signal we'd missed. The pain was real, we just had the wrong person.

That was enough to act on, it was a green light (read: go). We shifted the hypothesis and moved into a Listening Cycle — weekly interviews with a new target: controllers at $30–$100M SaaS firms — big enough to have the problem, but small enough to not build a custom solution. Thirteen interviews done to confirm what the BCS had surfaced. They didn't want another dashboard, they wanted their existing ERP to stop breaking. The positioning shifted from "cheaper processor" to "operational sanity." Already seeing higher retention, shorter sales cycles, wider reach with a more narrow kind of customer.

The company wasn't doing the wrong thing, the conversations they'd been having were exactly what a functioning business runs on — support, retention calls, closing plays. Those activities were doing what they were supposed to. They just weren't going under the surface. Nobody had separated the activity from the purpose.

That's the problem with "talking to customers" as a phrase. It covers everything under the sun — customer service answers what went wrong after the sale, NPS captures how someone feels at a moment in time, surveys collect opinions, success calls are working the retention angle, and a sales call is closing in motion. None of these are broken for being what they are. But when you use the same language for all of them, you can't see what's there and what’s not.

A sales call asks, "Will you buy this?" A discovery interview asks, "What are you already doing to solve this — and why?"

They produce different evidence. And when capital was cheap, the cost of conflating them was easy to absorb. A misaligned ICP, a stalled pipeline — you could run more outreach, throw more at the funnel, buy more ads. But that math has changed. Expensive money means the polite string-along you're getting from the wrong customer is now a liability you're actively carrying.

The reset for this team happened because the evidence got concrete enough to act on. Not a hunch. Customer signal, specific enough to name, specific enough to reposition around.

If you're sitting on a stalled pipeline and a team that swears they talk to customers, the question isn't whether they're having conversations. It's which job those conversations are doing. Right person, wrong seat.

And that's what the next issue is about: how to have the right one.

Thanks,
Skipper Chong Warson

_{Only humans buy anything, stop building for the wrong customer}

_—

_{If you want to know what a real discovery conversation looks like for your team, let’s talk: howthisworks.co/start}