We turned our worst support month into our best retention month. Here's exactly what we did.
Four months ago we had our worst support month since launching.
A feature update caused unexpected behaviour for about 30% of active users. Support volume tripled in 48 hours. Response times slipped. Some customers were waiting 6-8 hours for replies.
By every metric, it was a bad month.
By retention metrics one month later, it was our best month ever.
Here's what we did differently during that crisis:
Day 1: Acknowledge immediately, even without answers
We sent a personal email to every affected customer within 4 hours of identifying the problem. Not a fix — we didn't have one yet. Just: "We know what's happening, we're working on it, here's what we know so far, I'll update you by [time]."
Customers who got this email sent significantly fewer angry follow-ups.
Day 2-3: Over-communicate progress
We sent updates every 12 hours whether or not anything had changed. "Still working on it, here's where we are" is infinitely better than silence.
Day 4: Fix deployed. Personal follow-up to every affected customer.
Not mass email. Individual messages referencing their specific situation.
"Hi [Name] — the issue affecting your [specific feature] has been resolved. I checked your account and everything looks correct now. Let me know if you see anything unexpected."
Day 7: Proactive check-in
Reached out again. "Just checking in to make sure everything has been working well since the fix. We made some additional changes that should prevent this from happening again."
Outcome:
Of the ~30% of users affected:
- Churn rate: 2.1% (our normal monthly churn is 3.8%)
- NPS from affected users that month: higher than our average
Customers who were affected and handled well didn't just stay. They became the customers who wrote reviews and referred others.
The crisis wasn't a cost. It turned out to be the most efficient trust-building event we've had.
How do you handle product incidents from a support perspective? What's worked?
The proactive outreach before users even noticed the issue is the part that matters most here.
We're a two-person studio shipping SaaS tools and we've seen the same thing from the other side. The merchants and agencies we talk to don't churn because the product broke. They churn because they found out too late and felt ignored.
30% of users affected and you reached out individually to each one. That's brutal to execute but it's the only move that actually works at that stage. Templated "we're aware of the issue" emails don't cut it.
Did you automate any part of the individual outreach or was it fully manual for all affected users?
This is solid — especially the “over-communicate + personal follow-up” part.
One thing that stands out though:
You mentioned customers who were handled well ended up referring others — which is where things usually compound.
At that point, how people talk about you matters a lot. And that often ties back to how clear + memorable the brand is when they mention it to someone else.
Curious — have you seen people actually refer your product by name easily, or more like “that tool we use”?