Renewal Risk Assessment Template: How to Grade an Account 120 Days Out

What is a renewal risk assessment supposed to do?

Convert scattered account knowledge into a grade that a manager can compare across a book and a forecast can consume. Most renewal risk lives in a CSM's head as a feeling about an account. That feeling is often correct and completely unusable, because it cannot be rolled up, challenged, or trended.

The assessment below produces a letter grade and a required action. It runs first at 120 days before the renewal date, then refreshes at 90, 60, and 30. Enterprise accounts with formal procurement need an earlier start, usually 180 days, because by day 90 the budget conversation has already happened without you.

What signals belong on the scorecard?

Six, weighted toward outcomes rather than relationship warmth. Convert each score to points before weighting: 0 becomes 0 points, 1 becomes 50 points, 2 becomes 100 points. The weighted average of those points across the six signals produces a total from 0 to 100. Grade that total: A above 80, B from 60 to 79, C from 40 to 59, D below 40. The grade determines the intervention, not the probability.

Why does the support case pattern work in both directions?

Because zero cases and high case volume both indicate a customer in trouble, for opposite reasons. This is the signal most teams get backwards. A quiet account feels healthy. It usually means nobody is logging in.

In ORM customer data the pattern is consistent. Accounts with no support cases are at risk of churn. Accounts with seven or more cases in the last year are also at risk. Accounts with three to five tickets, typically tier 2 or tier 3 rather than severe, are less likely to churn. Those customers are engaged, getting help, and generally happy.

That makes the middle band the healthy one, which is why the scorecard scores it at 2 and scores both extremes at 0. It also means your support system is a churn data source, not only a service function, and the ticket volume field belongs in the same report as the usage trend.

What action should each grade trigger?

A specific intervention with an owner, escalating with the grade. The ten-day window matters. A grade with no deadline attached becomes a label, and labels do not change outcomes. The action is what the assessment exists to produce.

How do grades become a renewal forecast number?

Score accounts for two to three quarters, record what actually renewed, then use your own realized renewal rate for each grade band.

Skipping that calibration step is the common failure. Teams assign a probability to each grade on day one because the forecast needs a number, and those probabilities are guesses dressed as data. Two quarters of scored accounts against actual outcomes gives you real rates, and those rates are specific to your product and your segment mix.

Report the result inside the retention waterfall rather than as a standalone renewal number. ORM structures it monthly: beginning ARR, churned customer ARR, churned product ARR, product decrease ARR, new customer ARR, new product ARR, increased product ARR, ending ARR, where beginning ARR always equals the prior month's ending ARR. That reconciliation is where a renewal forecast stops being an opinion, because every dollar of movement has to land in a named bucket.

Both net revenue retention and gross revenue retention sit on that same waterfall. The risk grades explain the churn and contraction lines before they happen, which is the only useful time to explain them.

What makes a renewal risk process fall apart?

CSMs writing their own criteria. When each person defines risk differently, an A from one CSM and an A from another mean different things, and the roll-up is noise. RevOps owns the scoring rules and the definitions. The CSM owns the answers.

The second failure is running the assessment once. A grade produced at 120 days and never refreshed misses the sponsor who left in month two, which is the single change most likely to turn a B into a D.

The third is treating a low grade as a prediction rather than a task list. A D grade is not a forecast of churn. It is a statement that six specific things are wrong and somebody has ten days to start fixing them. Teams that hold that distinction see their forecast accuracy on renewals improve, because the assessment starts changing outcomes instead of describing them.