At-risk account playbook

Most churn is visible months ahead. This playbook covers how to spot risk early, diagnose the real problem, and run a focused save plan.

Quick answer — What is an at-risk account? A customer showing signals, such as declining usage, champion turnover, or negative sentiment, that suggest they may not renew.

Warning signals

Signal typeExamples
UsageDeclining logins, fewer active users, core feature drop-off
PeopleChampion leaves, executive sponsor goes quiet, new decision-maker
SupportRising tickets, repeated escalations, unresolved bugs
CommercialLate payments, budget cuts, procurement asks about alternatives
SentimentNegative survey scores, curt emails, canceled meetings
BusinessLayoffs, acquisition, strategy change at the customer

Step 1: triage

Not every yellow flag deserves the same response. Rank at-risk accounts by revenue at stake, time to renewal, and strength of the signal. A health score or AI risk model helps, but confirm with a human read of the account.

Step 2: diagnose the real cause

  • Ask open questions: "What would need to change for this to be a clear win for you?"
  • Separate symptoms (low usage) from causes (poor onboarding, wrong users, missing feature, budget shift).
  • Find the decision-maker, not only your champion.
  • Check history: was value ever demonstrated?

Root causes are catalogued in why customers churn.

Step 3: build a 30-day save plan

  1. Goal: what does "saved" mean (renewal, reduced scope, extended term)?
  2. Owner and sponsor: assign a lead and an internal executive if warranted.
  3. Actions: fix the problem, deliver a quick win, or reset onboarding.
  4. Stakeholder plan: reconnect with the sponsor and expand contacts. See multi-threading.
  5. Checkpoints: weekly reviews with a clear go/no-go.
  6. Commercial options: scope, term, or price only after value is addressed.

Escalation

  • Define when risk escalates to management, product, or executives.
  • Use internal reviews to unblock engineering or support.
  • Keep customer communication consistent: one voice, honest timelines.

Tone matters, especially under pressure. See handling difficult customers.

When the save fails

Run a proper post-mortem: reason, timeline, early signals missed, and whether a win-back is plausible later. See win-back. Record the reason so churn analysis gets better.

A one-page save plan template

FieldContent
Account / ARR / renewal date
Risk summaryWhat happened and when
Root cause hypothesisWhat we believe is driving risk
GoalWhat "saved" means
Actions (owner, date)Fixes, meetings, escalations
Stakeholders to engageSponsor, champion, new contacts
Executive support neededWho and for what
CheckpointsWeekly review, go/no-go date

What to say to the customer

  • Acknowledge the issue plainly.
  • Describe the plan and who owns it.
  • Give dates for updates, and keep them.
  • Avoid blaming the customer or other teams.
  • Ask what would make this a success from their view.

Escalation levels

LevelWhenWho joins
1Yellow signalsCSM and manager
2Red account with material ARRCS leader, support lead, product contact
3Strategic account at serious riskExecutive sponsor on both sides

Why most saves fail, and what to do differently

Saves tend to fail for predictable reasons. The team starts too late, fixes the symptom rather than the cause, or involves the wrong people. A discount offered to an unhappy customer who has stopped using the product just delays the same conversation by a year. The most reliable saves begin with a clear diagnosis, direct engagement with the economic buyer, and a visible change in the customer's experience within weeks.

It also helps to keep perspective. Not every account can or should be saved. Some customers were a poor fit from the start, and heroic efforts to retain them can drain the team. A good playbook includes criteria for deciding how much effort an account deserves, and a clean path for learning from the ones that leave.

Key takeaways

  • Diagnose before offering fixes or discounts.
  • Engage the economic buyer, not only the champion.
  • Show visible improvement within weeks.
  • Set effort levels by account value and fit.

Frequently asked questions

What is an at-risk account?
A customer showing signals, such as declining usage, champion turnover, or negative sentiment, that suggest they may not renew.
How do you save an at-risk customer?
Diagnose the real cause, assign an owner, deliver a quick win, reconnect with the sponsor, and address commercial terms only after value.
When should an account be escalated?
When revenue at stake is high, renewal is near, or the issue needs resources outside CS, per your defined criteria.
Can AI detect at-risk accounts?
It can flag patterns in usage, tickets, and conversations, but humans should validate and decide the response.
How early should you act on at-risk signals?
As soon as a credible signal appears. Most saves happen months before renewal, so waiting until the final weeks leaves few options.

See what CSMs really earn

Compare your pay against real CSM submissions — free, anonymous, takes 2 minutes.

B
Ben Hancock
Customer Success Manager · built this database to bring transparency to CS compensation