Upload a list, everything else is automatic
No manual entry — every field the model needs comes straight from your list. Nothing is left unfilled.
- Every weight comes from checking 16,359 real historical leads against what actually closed — not intuition.
- Hot ≈ 79% of similar leads have historically closed. Warm ≈ 49%. Cold ≈ 13%.
- A lead's score rises automatically once real engagement happens (a meeting gets booked, a call gets logged) — it isn't fixed at import.
Every imported lead, sorted by band
Everything from the last import, scored and banded. Click a card to open it in the walkthrough.
The rules, in full
Every gate and point value, plus a simulator that runs on leads already on your board — not a hypothetical.
Eligibility gates
Hard checks before any scoring. Fail one and the lead is filtered — never scored, never contacted.
| Gate | Rule |
|---|---|
| Size | Domestic <7 employees; Foreign HQ <3 (ESC/ESCP) or <7 (E-only) |
| Setup | No GST/entity/bank account on record (assumed true if not stated) |
| Source | Onsurity → routed to rollover queue, never scored fresh |
| Industry / workforce / proprietorship | Excluded industries, contractor-only workforce, unqualified proprietorships |
Points
Backtested against real Closed Won/Lost outcomes. See lead-score-v1.js for the full derivation.
| Signal | Group | Points |
|---|---|---|
| Employees 7–15 | Fit | +40 |
| Employees 16–50 | Fit | +15 |
| Employees 51–500 | Fit | −10 to −20 |
| Fresh lead | Fit | +25 |
| Rollover lead | Fit | −25 |
| Recently funded | Fit | +20 |
| Direct Traffic source | Fit | +15 |
| Non-brand Paid Search | Fit | −10 |
| 3+ sessions / visited pricing page | Fit (provisional) | +8 / +10 |
| SPOC-arranged meeting | Engagement | +20 |
| Client self-booked meeting | Engagement | +10 |
| Call logged | Engagement | +25 |
| Band | Score | Backtested win rate |
|---|---|---|
| Hot | ≥ 75 | 78.7% |
| Warm | 35–74 | 48.6% |
| Cold | < 35 | 12.6% |
Simulate an action
"What would happen if we did X to a group of leads we already have — before we actually do it." Pick a real segment from your imported board, pick an action, and see the projected before/after using the real band win rates above. Import a list first if you haven't.