6 early warning signs a customer is about to churn

The cancellation email is the last step. Weeks earlier, usage usually shifts: fewer logins, one feature instead of many, shorter visits. Nobody complained. The signs were in the data.

For customers who fade without tickets or feedback, see silent churn. For how those signs unfold over time, see how customers quietly disengage.

Leading vs lagging indicators

Lagging indicators tell you what already happened: churn rate, revenue lost, survey scores after the fact. Useful for reporting, too late to save that account.

Leading indicators tell you what is starting: login drops, narrower usage, stalled onboarding. You can still reach out.

Most teams review lagging metrics monthly. Leading indicators need a weekly or daily look at trends while you can still act.

6 early warning signs

Six patterns predict churn across most SaaS products. Scan the table, then use the checklist for what to do.

1. Login frequency drops

What it looks like: A daily user becomes weekly. A weekly user becomes every other week. The count may still look acceptable; the trend is the warning.

What to do: Send a short, specific nudge tied to value they used to get: "Your team ran reports every Monday last quarter. Want a quick walkthrough of what's new?"

2. Feature usage narrows

What it looks like: They used to explore. Now they log in, do one thing, and leave. One workflow holds the whole relationship.

What to do: Highlight one adjacent feature that supports their existing workflow. Do not pitch the whole product. Show how the next feature saves time on what they already do.

3. Session duration shortens

What it looks like: Each visit does less. Fewer pages, fewer actions, less time in the product. They are showing up but accomplishing less.

What to do: Check for friction: broken workflow, missing integration, or a champion who left. Offer a 15-minute session to unblock the one task they still run.

4. Team seats go inactive

What it looks like: One quiet user is a concern. Three quiet seats on the same account is organizational disengagement.

What to do: Identify who still logs in and who stopped. Reach the active contact with seat-level data: "Four of six licensed users have not logged in this month."

5. Support tickets spike then stop

What it looks like: A burst of tickets, then silence. They tried to get help, then stopped asking. No tickets is not always healthy.

What to do: Reopen the thread personally. Reference their last ticket and ask if the issue was resolved or if they gave up. A human follow-up beats another automated survey.

6. No milestone progress

What it looks like: Onboarding never finished, key features never activated, or usage never expanded beyond day one. They never reached full value.

What to do: Reset to one milestone, not the whole onboarding program. "Let's get [one outcome] live this week" beats a generic training invite.

How these signs connect

The signs rarely appear all at once. Gaps between visits often widen first. Then logins and session depth fall. Milestones stall last. That order is why early detection matters: you can act when only one or two signs are visible.

For the full five-stage timeline (thriving through gone), read how customers quietly disengage before they cancel.

When multiple signs show up

Match response to how many patterns are active.

One sign (monitor range, roughly 40-69): Light outreach. Tips, use-case content, a helpful nudge. Save rates are often 60-80% if you catch it here.

Two or more signs (at-risk, roughly 40-69): Personalized outreach. Reference the specific behavior: unused features, longer gaps, stalled setup. Save rates often 30-50%.

Most signs declining (critical, roughly 0-39): Human conversation. Call, video, or executive check-in. Automation alone is usually too late. Save rates often 10-20%.

Account nearly dark (churning, roughly 0-39): Last-resort save. Acknowledge the lapse and offer one clear path back. Win-back success is often 5-10%.

Acting on the first sign is far cheaper than waiting for four. For building proactive habits across the team, see proactive vs reactive customer success. To roll these signs into one score, see customer health score.