Configure propensity scoring and signals

Propensity scoring ranks accounts by how likely they are to be worth working now. GTM Engine researches the buying signals your team defines, scores each account, and writes the reasoning back to the account so sellers can prioritize and personalize outreach.

What You'll Build

You will create at least one research signal, test it against an account, and turn on automatic scoring for new accounts.

Estimated time: 10-20 minutes.

Prerequisites

• You are an Admin or RevOps user.
• Your products and ICPs are configured in Settings → Products.
• You have at least one account to test against.

1. Open Propensity Scoring Config

Go to Generate Pipeline → Propensity Scoring Config.

This page controls the signals GTM Engine uses when it scores accounts. Top Target Accounts uses those scores to rank accounts and explain why each one is worth working.

2. Add a Research Signal

Click Add Signal and define the signal you want GTM Engine to research.

Good first signals are specific buying indicators, such as:

• Recent funding or budget expansion.
• Hiring for roles related to your product.
• Technology usage that suggests fit.
• Public initiatives that match your use case.
• Competitor usage or migration signals.

Each signal should answer one clear question. Avoid broad prompts like "is this a good company?" because they produce generic scores.

3. Review or Generate Signal Instructions

For each signal, confirm the research and scoring instructions:

• What evidence should GTM Engine look for?
• What should count as a strong, medium, or weak signal?
• What should happen when public evidence is missing?
• Should the signal apply only to certain products, ICPs, or segments?

Use concrete scoring rules so different accounts are evaluated consistently.

4. Test the Signal

Run a test against a representative account before enabling automatic scoring.

Check that:

1. The research uses relevant evidence.
2. The score matches your team's judgment.
3. The reasoning is useful for outreach.
4. Missing evidence is handled honestly instead of invented.

If the output is too generic, tighten the signal description or scoring instructions.

5. Publish and Enable Automatic Scoring

Save or publish your signal changes, then turn on automatic propensity scoring.

When enabled, GTM Engine scores new accounts added directly or synced from your CRM. Existing accounts can still be scored manually or in bulk from Records → Accounts by running the propensity scoring workflow.

6. Set a Refresh Cadence

Choose how often GTM Engine should refresh existing propensity scores.

The refresh cadence is optional. When set, GTM Engine checks once per day for already-scored accounts whose last propensity run is at least that many days old, then queues a new propensity scoring workflow for each account.

Use a shorter cadence, such as 7 or 14 days, when your market signals change quickly. Use 30, 60, or 90 days when accounts are stable or you want to conserve credits.

7. Use Scores in Top Target Accounts

Open Generate Pipeline → Top Target Accounts to review accounts ranked by propensity.

From there, you can:

• Sort and filter by propensity score.
• Open the score reasoning and signal-level results.
• Find buyer-persona contacts for high-fit accounts.
• Enroll selected accounts into outreach campaigns.

Success Check

You are done when:

• At least one signal is configured.
• A test account produces useful signal research and scoring.
• Automatic propensity scoring is enabled.
• Refresh cadence is either off intentionally or set to the period your team wants.
• New or scored accounts show propensity values in Top Target Accounts.

Common Pitfalls

• Enabling scoring before testing one account.
• Creating too many overlapping signals.
• Using vague signals that do not map to a real buying indicator.
• Forgetting to publish signal changes before expecting automation to use them.
• Expecting old accounts to score automatically without running a backfill or bulk workflow.

Next Steps

• Run a workflow over many records
• Import and enrich a list of prospects
• Build an outbound prospecting workflow