Automating Lead Scoring
Scoring runs automatically when a call completes. This guide covers what happens on that path, and how to build automation on top of it.
The Automatic Path
When a call ends, the platform evaluates the transcript and captured fields, produces a lead score with a breakdown and a stated reason, and flags leads that meet the hot-lead criteria. No configuration is required for this to happen.
Reading the Score Breakdown
Each score comes with a breakdown and a reason. Read these before trusting or acting on scores in bulk — the reason tells you which signals drove the result, which is what you need in order to judge whether the score is behaving sensibly for your business.
💡 Tip: Review the reasons on twenty or thirty scored calls before you build automation on top of scoring. Automation amplifies whatever the scoring is doing, including its mistakes.
Triggering Off Scores
Use the lead scored trigger rather than the inbound-call trigger. It fires after scoring completes, so the score and its breakdown are available to your conditions. Triggering on the call itself risks evaluating conditions against a score that does not exist yet.
A Practical Tiered Setup
- Hot leads — immediate SMS to whoever can call back fastest
- Warm leads — email with the transcript, handled within the working day
- Low-scoring calls — no alert; they appear in the daily digest only
⚠ Note: The point of tiering is protecting attention. If every call generates an alert, the tiers are decorative and the hot-lead alert loses its meaning.
When Scoring Fails
Scoring can be skipped or fail — very short calls, unusable audio, or calls that end before anything meaningful is said. Build workflows so a missing score does not silently drop a lead: if your only path to a human requires a high score, an unscored call disappears.
✓ Success: A good test: take a call with no score at all and confirm someone would still see it. If not, add a catch-all path.