Three months ago, I had no idea users were about to cancel. I'd just wake up to a Stripe notification and stare at it. Another one gone. No warning. No reason. Just $79 or $149 quietly disappearing.
The worst part wasn't losing the revenue. It was that I had no idea it was coming.
I tried everything the standard playbook says. NPS surveys nobody fills out. Win-back emails are sent three days after someone has already cancelled and moved on. Exit surveys with five dropdown options where users just click the first thing and leave. I was collecting noise and calling it data.
Then I stumbled onto something that changed how I think about churn entirely.
The real problem isn't that users cancel. It's that by the time they cancel, they have decided weeks ago. The cancel button is just the last formality. The actual decision happened somewhere in the middle of your product, quietly, when nobody was watching.
So I started using Flidget. Two things happen with it.
First, it scores every user as Healthy, Risky, or Drifting based on actual behaviour. Login frequency, feature usage, and session patterns. Not a vanity metric. A plain-English reason next to each user, telling me exactly what's slipping and when to act.
Second, when someone does hit cancel, a short chat opens right there on the page. Not a form. Not a dropdown. An actual conversation. They type or speak for ten seconds, and I get their real reason, tagged in my dashboard.
Here's what changed specifically this month.
My dashboard flagged 8 users as Drifting. I reached out to 6 of them with a personal email before they ever saw my cancel page. Not a campaign. Just a short, honest note to say I noticed they hadn't been back and to ask if something got in the way.
Three replied. Two had a specific problem I fixed in an afternoon. One needed to pause for a month, which I set up in two minutes.
That's $1,400 in MRR I kept without any of them reaching the cancel button.
The other $600 came from two users who tried to cancel. The exit chat caught them. One mentioned a competitor. One mentioned CSV export. I replied to both personally within an hour. One stayed. One left but told me exactly what to build.
Total this month: $2,000 in MRR that would have silently disappeared is still in my Stripe dashboard.
The thing nobody tells you about churn is that most of it isn't dramatic. Users don't rage quit. They just slowly stop caring, and one day click cancel while doing a subscription audit. By then, you've already lost them.
The only way to win is to know before they've decided.
If you're still finding out about churn from Stripe notifications, you're already too late. Please be in the room when the decision is being made, not after the paperwork is signed.
Flidget.com is free to start. One script tag. I was living under an hour.
Happy to answer questions about the setup or what the drift scores actually look like in practice.