I ignored churn for years and it quietly capped every SaaS I built

I've launched a handful of SaaS products. Most flopped. The ones that got traction all taught me the same painful lesson, and I wish I'd learned it earlier: I was obsessed with getting new customers and completely blind to the ones quietly leaving out the back door. I'd celebrate signups while roughly the same number churned, running up a down escalator without realizing it.

When I finally dug into the churn data, two things genuinely surprised me:

1. By the time a cancellation shows up in your revenue number, the customer decided to leave weeks earlier. Churn is a lagging indicator — the warning signs (usage dropping, a key user going quiet, a failed payment) are all there before the cancel, if you're looking at leading indicators instead of last month's MRR.

2. A big chunk of "churn" isn't people choosing to leave at all — often 20–40% is cards that silently failed and dunning never recovered. The customer never decided anything; the payment just quietly died. And it's usually the easiest revenue to win back, because they weren't even unhappy.

That second one broke my brain a little. Most dashboards lump both together as one "churn" number, so you end up fixing the wrong problem.

I got obsessed enough with this that I ended up building a tool around it (Exeechain), but honestly the lesson stands with or without any software: retention is a leading-indicator game, and most of us only watch the lagging number.

Curious how other founders here handle it: do you separate voluntary churn from failed-payment churn? And have you found any signal that reliably tells you a customer's about to leave before they do?

Happy to share more of what I found in the data if it's useful — and if anyone wants their own churn broken down (voluntary vs failed-payment split), I'm genuinely happy to do a few for free.