Skip to content

New in October

Core concepts

Scores and the quality threshold

How the score threshold decides which leads reach you, how to tune it, and how to read the run funnel.

Last updated 2026-07-14

Discovery only surfaces leads whose score clears your threshold. It's the main quality dial you control per run.

How it works

Each run has a threshold (default 60, adjustable from 0 to 100). Scored candidates at or above it are ranked and the best ones — up to your target count — become suggestions. Everyone else is kept visible as data, never silently discarded.

Reading the run funnel

Every run shows its funnel: how many profiles were found, deduplicated, dropped by your "Don't target" exclusions, filtered by location, scored, enriched, and finally suggested. Two parts are worth a look when results feel off:

  • Near misses — the strongest candidates that didn't clear the bar, with their scores and stage. If they look good to you, your threshold may be too strict or your ICP too narrow.
  • Exclusion count — how many candidates your negative filter removed, with examples of what matched.

Choosing a threshold

Higher thresholds mean fewer, better leads; lower means more volume with more noise. Three practical rules:

  1. Start at the default and read the reasoning on a run's suggestions — adjust based on what you see, not on the number alone.
  2. If a run returns very few leads, check the near misses before lowering the bar: sometimes the fix is a broader ICP, not a lower standard.
  3. Remember the score is fit only — timing ("why now") is a separate signal shown on the lead, so a 75 with a hot timing signal can be more valuable than a cold 85.