Skip to main content
Lead scoring runs after enrichment to give every lead a clear priority tier. Scores drive automatic routing: hot leads get a Striker task, warm leads enter nurture, cold leads are archived.

Scoring scale

ScoreTierAutomatic action
8–10HotStriker outreach task created immediately
5–7WarmEnter 4-email nurture sequence
1–4ColdArchived, no outreach

Scoring criteria

ICP fit (40%)
  • Company size match
  • Industry match
  • Funding stage match
  • Geography match
Signal strength (30%)
  • Recent trigger events (funding, hires, tech changes)
  • Time since trigger (fresher = higher score)
  • Number of concurrent signals
Contact quality (20%)
  • Email verification status (valid > catch-all)
  • Title seniority (C-level > VP > Director > Manager)
  • LinkedIn profile completeness
Budget indicators (10%)
  • Job postings signaling relevant spend
  • Tech stack choices that imply budget
  • Press coverage and growth signals

Configuration

SettingValue
ModelFast (Haiku-class)
Max turns6
ExecutionInline
Toolslookup_leads, search_memories, update_lead
Lead scoring uses the fast model because it’s pattern matching against defined criteria — not complex reasoning. This keeps scoring fast and cheap across large lists.

How to invoke

Score the 15 leads Specter found yesterday.
Re-score our warm leads list with the updated ICP I just set.