The customers you're not talking to (but should be)
Stop surveying happy customers. Start calling lost ones.
Hi there,
A few years ago, I was digging into churn for a subscription brand. The cancellation survey data was… fine.
“Too expensive” topped the list. “Not using it enough” came second.
Useful? Sort of.
Actionable? Not really.
So I did something that felt awkward at the time:
I emailed 50 people who’d cancelled in the previous month and asked if they’d be open to a 15-minute call.
Eight said yes and those eight conversations taught me more than months of survey data ever had:
One woman told me she loved the app — but only used it during particularly stressful periods. A monthly subscription simply didn’t make sense for her life.
Another explained that “too expensive” actually meant, “I couldn’t justify it to my husband,” which is a completely different problem to solve.
This is the pattern I see again and again: cancelled customers will tell you things your active customers won’t.
They have nothing to lose.
The relationship is already over.
And that honesty is exactly what you need.
Why churned customers are your best research subjects
1. They’ve made a real decision
Active customers are still in the relationship. They’ll tell you what they like, what they might want, what sounds good in theory.
Cancelled customers have actually weighed your product against alternatives and chosen to leave. That decision, the moment they clicked “cancel,” contains more truth than any hypothetical preference.
2. They are more critical
Happy customers adapt. They work around limitations, forget early frustrations, and normalise the experience.
Churned customers remember exactly what friction felt like. They can tell you the moment it stopped working, because that moment made them act.
3. They’ll be honest in ways that feel uncomfortable
Active customers want to be nice; they like your brand and want you to like them. They’ll soften feedback, hedge criticism, and tell you what they think you want to hear.
Someone who’s already left has no reason to protect your feelings. That bluntness is a gift.
I feel like a lifetime of being Dutch has made me strangely enjoy the bluntness of cancelled customer feedback.
The questions that actually reveal something
Most cancellation surveys ask, “Why did you leave?” and offer a multiple-choice list of options. You get data, but not insight.
The questions that actually work go deeper: what happened, when it happened, and what might have changed the outcome.
Start with the trigger:
What led you to stop using [Brand]?
What made you cancel on that specific day?
That second question matters. There’s usually a gap between “thinking about cancelling” and “actually cancelling.”
Something closed that gap… a frustrating experience, a charge they noticed, a competitor ad. That trigger tells you where you lost them.
Then follow the thread based on what they say:
Reason 1: Price
What would have made it feel worth the price?
Was there something you expected it to do that it didn’t?
“Too expensive” is never the whole story. It usually means “too expensive for what I was getting.” The follow-up reveals what value they were missing.
Reason 2: Not using it enough
If they weren’t using it enough:
How often were you actually using it?
What would have led you to use it more?
Non-usage is a symptom, not a cause. Maybe the product didn’t fit their routine. Maybe they forgot it existed. Maybe it solved a problem they don’t have often enough.
Reason 3: Missed certain features
If they wanted features you don’t have:
What features were most important in choosing a product like this?
Did you think we had that feature, or is this a new need?
This tells you whether it’s a positioning problem (they expected something you don’t offer) or a product gap (they need something you haven’t built).
Reason 4: Short-term need
If they only needed it short-term:
Were you satisfied with using [Brand] for that purpose?
What might bring you back in the future?
Not all churn is bad churn.
But understanding the short-term use case might reveal a segment you’re accidentally attracting — or one you could intentionally serve differently.
Reason 5: Switched to a competitor
If they switched to a competitor:
What specifically led you to switch to [competitor]?
Don’t be afraid to ask this directly.
The answer reveals how they perceive your competitive positioning, and it’s often different from how you see it.
Always end with:
What are you using now instead?
What led you to choose them?
Even if they didn’t switch to a direct competitor, they’re solving the problem somehow.
Understanding that alternative tells you what you’re really competing against.
How to actually get them on a call
This is where most founders get stuck. “They’ve already left. Why would they talk to me?”
More people will say yes than you expect. Here’s what actually works:
1. Email personally, not from a marketing address
Send it from your own name, not “The [Brand] Team.” Keep it short. Something like:
“Hi [Name], I saw you recently cancelled your subscription. No pitch, I’m just trying to understand what we could have done better. Would you have 15 minutes for a quick call? I’d really appreciate the honesty.”
2. Offer a small incentive (but don’t lead with it)
A £10 gift card or a free month if they ever return can help. But lead with the ask, not the bribe. People respond to genuine requests for help more than you’d think.
Also, make sure the incentive isn’t just your product/app; they left for a reason.
3. Time it right
Reach out within 2-4 weeks of cancellation.
Too soon, and they’re still annoyed.
Too late, and they’ve forgotten the details.
4. Accept that most won’t respond
That’s fine. You don’t need fifty calls. 5-10 honest conversations will show you patterns. If you’re not seeing patterns after eight calls, keep going until you do.
5. Don’t try to win them back on the call
This is research, not sales. The moment you pitch, they’ll stop being honest. Thank them, learn from them, and let them go. Don’t ever try to excuse your flaws.
What to do with what you learn
A few calls won’t give you statistical significance. That’s not the point.
What they will give you:
Language: How real customers describe the problem (use this in your copy)
Triggers: What actually caused the cancellation (not the rationalisation)
Gaps: What they expected vs. what they got
Alternatives: What they’re using instead and why
After five or six calls, you’ll start hearing the same things. That’s your signal. Those patterns are more valuable than any survey percentage.
Recommendation
In every edition of Growth Waves, I also share a resource to check out related to the week’s topic.
If you’re not already doing cancellation interviews, start this week. You don’t need a system: just email at least 20 recent cancellations personally and see who replies.
I’d also recommend reading The Mom Test by Rob Fitzpatrick if you haven’t. It’s technically about customer development for new products, but the principles of asking good questions (and not leading the witness) apply perfectly to churn conversations.
Your cancelled customers know something you don’t.
They know exactly when your product/app stopped being worth it to them, and what might have changed that.
All you have to do is ask.
Till next time,
Daphne