Set up renewal and expansion scoring

You want to know which of your customers are heading toward a renewal you can grow. This is the guide for that goal, and it is the one most likely to tell you your CRM cannot answer it yet.

What it represents

Renewal / expansion is the fourth of the four cards on Create a Score. It ranks current customers by how likely they are to renew and expand, learned from the renewals your book has already won.

It shares its population with At-risk customers and points at the opposite outcome. That pairing is deliberate: both questions are asked of the same accounts, and the difference is which end of the renewal you are looking at. If the stated goal on your account is renewals, both cards are relevant and their difference is one line each.

Which to build first is a workflow question, not a modeling one. An at-risk model directs retention effort. This one directs growth effort. Teams with a named CSM function usually want the risk model first, because the cost of missing a churn is higher than the cost of missing an upsell.

How it’s calculated

The population line reads Trains on: accounts that are current customers, the same filter the at-risk card carries · the account type field on Salesforce, the lifecycle stage on HubSpot. You can only renew somebody who is already a customer, so the population is not an optional narrowing here, it is part of the question.

Success is the won renewal · Renewal - Won on a standard Salesforce org, the generic won stage (closedwon) on HubSpot, where the preset has no renewal typing to point at · dated by the deal’s close date.

That definition depends on something many CRMs simply do not carry: a way to tell a renewal apart from any other deal. Where renewal typing is absent, this target has no honest reading. It would either find no positives at all, or it would quietly become a general won-business model on a customer population, which is a different question wearing this one’s name. The preflight checks for this before you train rather than after, and it refuses rather than degrades.

Set it up

1. Pick the Renewal / expansion card

   Fourth of the four at the top of Create a Score. It sits next to At-risk customers on purpose, so the two renewal-side questions are visible together with their difference stated in one line each.

   If your org answered renewals as its scoring goal during onboarding, neither card is pushed at you ahead of the other, because that goal genuinely spans both and choosing for you would be a guess.

2. Read the Trains on line

   Trains on: accounts that are current customers, on the card and again above Success definition. Same filter as the at-risk card, for the same reason.

   Read it against your own customer marking before anything else. If it does not resolve, this card and the at-risk card fail together, and they fail for the same fixable reason.

3. Confirm Connection and Primary entity

   Primary entity arrives set to Accounts. The record being ranked is the customer, and the renewal deals under it are what the label is read from.

   Leave it. An expansion question asked of deals rather than accounts is a different score, and the deal-side version of it is Won business.

4. Check the success criteria against your renewal stages

   Success criteria holds the won-renewal condition. This is the control to read most carefully on this card, because it names two things at once: the stage that means won, and the fact that the deal was a renewal rather than new business.

   If your org runs renewals through the same stages as new business with nothing distinguishing them, no edit here will fix that. What you need is a renewal type or a renewal pipeline in the CRM, and the preflight will name exactly what it looked for.

5. Confirm the Success date field

   Success date field is the close date on the renewal deal, and it anchors the lookback window the same way the at-risk model does. An account with no resolvable renewal date is counted as unresolved rather than filled in.

   Change it only to another field that is written at the moment the renewal resolves.

6. Leave the readout on More likely to convert

   What does a high score mean here? stays on More likely to convert. A high renewal and expansion score is upside, so the bands read hot · warm · cold and a high score reads as opportunity.

   This is the one place the two renewal cards diverge in the readout. Same accounts, same window, opposite polarity. If you find yourself wanting More at risk on this card, the score you want is the at-risk card.

7. Press Create score and read the preflight

   Checking your data... runs the population count, the success-date check, and the renewal-typing check. On this card the third one is the gate.

   Where the connection's fields have not been read yet, renewal typing comes back as a note rather than a pass, saying it will be checked again at train time and naming what it looks for. That is honest rather than evasive: an unread schema is not evidence of a missing field.

If the preflight refuses

We did not find renewal typing in this connection is the refusal this card exists to surface early. The block names what it searched for · a renewal type on the deal object, or a renewal stage on the stage field · so the finding is actionable inside your CRM rather than inside our form. The exit offered is Use Won business instead, because that question runs on the deal outcomes you already have. Take it if what you actually want is a ranked pipeline. Do not take it as a substitute for renewal scoring, because a won-business model on a customer population is a different claim.

No records match this cohort means the customer filter matched nothing. The exit is again Use Won business instead, since open pipeline needs no customer flag, and Include every record instead is offered when a population filter is present. Clearing it on this target means training on accounts that have never bought, whose renewal history is empty by definition, so it is rarely the right move here even though it is the right move on some other cards.

No success date field blocks with Continue in Advanced as its exit, carrying forward the object, the population, and the criteria that already resolved.

A warning about a thin population never blocks. Under 30 records the count is shown and travels with the result, because a rate read off a handful of renewals is a number nobody should act on.

What it means for you

The score is a rank of your customers by renewal and expansion likelihood, from 0 to 100. It is not a revenue forecast and it does not estimate an amount · it orders accounts, and the amount is a question your CRM already answers.

Read it as the mirror of the at-risk model rather than as its replacement. Where both exist, the interesting accounts are the ones the two disagree about, and the routine ones are the ones they agree on. An account that is high on both is usually a data finding: something in its history is loud enough to look like both outcomes, which normally means a field that records the renewal decision itself has reached the model.

One honesty note about the sample workspace. The renewal and churn shape in the demo data is synthesized rather than measured, so a persuasive result there is a statement about the generator and not about this model. Judge this one on your own book.

When Advanced is the right door. Take Start from scratch (custom) when renewal in your business is not a deal at all · a contract record, a subscription object, a billing system field · or when expansion means a quantity change rather than a stage change. The criteria builder can express targets on objects the cards do not touch, and it asks you the population question too, so a custom score means you defined who is included.

Editing this card rather than leaving it keeps its lineage. Change the criteria and the score records that it diverges · the create flow says so in one line, “You changed what counts as success from the Renewal / expansion starting point.”, and the saved score carries the equivalent note. It never blocks. Keep my version retires the note for good once you confirm the change was intentional, and the note returns only if the divergence set changes or the starting point’s own definition moves under you.

Check your understanding

Your connection carries renewal opportunities typed as renewals, and 220 accounts are marked as customers. You pick Renewal / expansion, the preflight passes with 220 records in this cohort and a renewal-typing note, and the first run reports a 70% base rate.

Read that back. The 220 is the customer filter resolving, and it is the reason no prospect is in the ranking. The 70% is the share of those renewals your book won, which is a high base rate and a legitimate one · most renewals renew. Because the readout is set to More likely to convert, a high score reads as upside and the bands say hot rather than at risk. Point the same population at the lost-renewal stage instead and you have the at-risk model, with the same 220 accounts, the same window, and every word around the number inverted.