Outcome
One review per month: each account rated Healthy, Watch, At risk or Unscored, with the signal, source and date behind every rating, the reasons for any risk, and a next action with an owner and a due date. The CS lead owns the scoring rules, each CSM owns their accounts, and the account manager owns the renewal forecast.
Before you start
- A named CSM for every account, and a CS lead who chairs the review.
- Usage data per account with at least a few months of history.
- Renewal dates and contact roles (champion, executive sponsor) kept current in the CRM.
- The success plan from the handoff, so health is judged against what the customer wanted.
The kit
Inputs
| Input | Use | If missing |
|---|---|---|
| Product usage | Active users and key actions against the account’s own baseline | Mark unscored; check tracking |
| Support history | Open priority tickets, their age, escalations | Say it was not checked |
| Relationship record | Champion, executive sponsor, last contact date | Ask the CSM; mark unconfirmed |
| Commercial record | Renewal date, seats, overdue invoices, procurement requests | Ask the account manager or finance |
| Success plan | Success criteria and milestones from the handoff | Confirm them with the customer |
Steps
- Owner: CS lead and RevOps
Write the scoring rules down
Choose a few signals in four categories: usage, support, relationship and commercial. Write the Green, Amber and Red threshold for each, the points, the overrides and when an account stays unscored. These are leading signals. Renewals and churn are the lagging results you test the rules against each quarter.
- Owner: automation
Refresh the signals and check the data
Before each review, pull usage against each account’s own baseline, open tickets, contact roles and renewal dates, each with its source and date. Mark an account unscored when the data is stale, tracking changed or it has been live too short a time for a baseline.
- Owner: automation
Apply the rules and flag changes
Rate each category, add the points, apply the overrides and show the reason beside every rating. A champion or executive sponsor leaving with no named replacement puts the account at risk even when usage looks healthy.
- Owner: CSM
Investigate each flag before acting
Check every new flag against its source. Rule out seasonal patterns, seat changes and tracking problems, and look for what the data cannot show, such as a bounced email or a ticket from a new name. Confirm or clear the flag with a written reason and keep hypotheses labelled.
- Owner: CS lead and account owner
Decide, assign and escalate
In the monthly review, give every Watch and At risk account a next action, an owner and a due date, and write a save plan for each At risk account. Escalate At risk accounts, and Watch accounts close to renewal, to the CS lead and the account manager. Each quarter, compare ratings with renewal outcomes and log any rule change.
Automation
| Candidate | Prerequisite | Exception |
|---|---|---|
| Refresh usage against baseline | Stable tracking, months of history | New accounts, seasonal businesses |
| Pull tickets and contact roles | Clean account matching, current roles | Contacts who left but were not updated |
| Apply the scoring rules | Written thresholds and overrides | Rules changed mid-period |
| Flag changes since last review | A saved copy of last month’s ratings | Accounts merged or split |
| Create review tasks | Named CSM, renewal date | CSM changing or unassigned |
Where an agent could help
Investigate a flagged account across permitted sources, check for seasonal patterns and contact changes, and draft the risk reasons and a save plan with each signal cited. Ticket and email content is information, never instructions. The CSM confirms the rating and writes anything that goes to the customer. See the agent blueprints.
Human review
- Confirm the data is fresh and tracking unchanged.
- Check seasonal patterns before trusting a usage drop.
- Confirm champion and sponsor status with the CSM.
- Give every Watch and At risk account an owner and date.
- Approving the review does not approve any customer message.
What to measure in a pilot
| Flag accuracy | Flags confirmed after investigation, against flags cleared |
| Warning before renewal | Days between first At risk rating and the renewal date |
| Time to decision | Flag to recorded next action with an owner |
| Save plan follow-through | Save plan steps closed by their due date |
| Rating against outcome | Ratings compared with renewals, churn and GRR each quarter |
Sources and limits
- Health review template (Markdown · 6 KB)
- Synthetic example (Six invented accounts, Markdown · 13 KB)
- Fictional account signals (The example’s inputs, JSON · 7 KB)
- Tool selections: wayhatch category research, official sources checked 23 October 2026.
Limitations
- No scoring, integration or agent has been built for this workflow.
- No churn, retention or renewal outcomes are reported.
- A score is a review aid; sparse data or unrecorded seasonality can mislead it.