How to Interview Your Churned Customers (Without Making It Weird)

Why most founders skip this and shouldn't

You lost a customer. They unsubscribed quietly, or sent a one-line email ("not the right fit right now"), or just let the card fail. You feel a little embarrassed, a little defensive, and you tell yourself the reason is probably price or "they weren't our ICP anyway."

It usually isn't.

Talking to churned customers is the single highest-leverage research activity for an indie B2B SaaS founder. Five 25-minute conversations will teach you more than three months of staring at Mixpanel. The reason most founders don't do it: it feels awkward to ask someone who just left to spend half an hour explaining why.

Here's a 5-step playbook for getting those conversations — and the one question that actually surfaces real reasons instead of polite ones.

Step 1: Reach out within 7 days, and be honest about why

The window matters. After two weeks they've forgotten the specifics. After a month they're confabulating a tidy narrative that may or may not match reality.

Send a plain-text email from your founder address. Not from "team@" or a tool that adds tracking pixels. Something like:

Subject: A favor (and no, I'm not trying to win you back)

Hey [name] — I saw you canceled [Product] last week. I'm the founder. I'm not going to pitch you on coming back. I'm trying to figure out what we're getting wrong, and a 20-minute call with someone who actually left is worth more to me than a hundred surveys from active users. Would you be open to it this week or next? I'll send a $50 Amazon card as a thank-you either way.

Three things this does:

• Removes the sales fear ("I'm not going to pitch you").
• Flatters the specific person rather than treating them as a data point.
• Compensates them honestly. $50 is enough to feel respectful without feeling like a bribe.

Expect roughly 1 in 4 to say yes. If you got 20 cancellations last month, that's 5 calls — exactly what you need.

Step 2: Don't prep too hard

You'll be tempted to build a 15-question script. Don't. Long scripts produce shallow answers because you're rushing to get through them, and the interviewee can tell.

Bring six anchor questions, max:

1. Walk me through the day you decided to cancel. What were you doing?
2. What was happening in your work the week before that?
3. What did you replace us with — another tool, a spreadsheet, nothing?
4. When you first signed up, what did you hope would change?
5. At what point did you realize that wasn't going to happen?
6. If we'd called you a month before you canceled, what could we have done?

Notice these are mostly questions about their world, not your product. That's the point.

Step 3: Shut up

This is the hardest part. In your first few interviews you will:

• Explain features they apparently missed.
• Defend pricing decisions.
• Promise that thing on the roadmap.
• Fill silences after 3 seconds because they make you anxious.

Don't. Every second you're talking, you're not learning. The rule I use: after they finish an answer, count to four before you respond. They will almost always keep going, and the second half of the answer is where the real stuff lives.

If you must say something, say "tell me more about that" or "what do you mean by [exact word they used]?" That's it. Those are your two phrases.

Step 4: Ask the one question that surfaces the real reason

Most churn reasons people give you are second-order. "Too expensive" usually means "I didn't get enough value to justify it." "Missing feature X" usually means "I couldn't get the outcome I came for, and X is what I think would have fixed it."

The question that cuts through:

"If [Product] had been completely free, would you still have canceled?"

Then wait.

If they say yes, price was never the issue — you have a value or fit problem. If they hesitate, you're learning where the actual line is. If they say no, but then qualify it ("well, if it were free I'd probably keep it around just in case"), that's even more useful: it tells you they didn't trust it enough to depend on it.

Follow up with: "What would have had to be true for you to keep paying full price?" Now you're getting roadmap.

Step 5: Write it up the same day, in their words

Within 2 hours of the call, dump your notes into a doc. Use their exact phrasing — not your sanitized version. "It felt like another tab I had to remember to open" is worth ten times more than your translation ("low engagement / habit formation issue").

Keep a running file. After 5 interviews you'll start seeing the same phrases. After 10 you'll see them in different words from different people, which is when you know it's signal and not one person's hot take.

What to actually do with all this

Here's where most founders get stuck. You've done 5 interviews. You have 40 pages of notes. You have a vague sense that "people don't get to value fast enough" but no clear next move.

The work now is synthesis: clustering the quotes, ranking the themes by frequency and severity, mapping them to fixes you could ship, and figuring out which fix touches the most churn risk. Doing this by hand in a Notion doc is fine — slow, but fine. It's also where the analysis usually dies, because it's tedious and there's always a more urgent fire.

This is the part Pulse was built for. You drop in your interview notes (or transcripts), and it clusters the themes, pulls the verbatim quotes that back each one, and suggests which problems map to roadmap moves vs. onboarding fixes vs. pricing changes. Five conversations become a ranked list of bets, instead of 40 pages you'll never reread.

But the interviews themselves? Those are on you. No tool replaces a 20-minute honest conversation with someone who left. Go send the emails.