Policy
How the weekly Sales & CS queue decides — and how it compares to a naive combined score. Adjust the two levers below and watch the queue recompute.
Live policy and capacity controls
You can change only two things, both allowlisted: the acquisition intent minimum (20–70, step 5) and the total weekly human-action capacity (5–30). Changing capacity reprioritizes the queue — it never changes whether an account qualifies; qualified overflow stays visibly deferred.
A prospect needs buying intent at least this high to qualify for sales-assist (20–70, step 5).
Total Sales/CS actions this week (5–30). Capacity changes prioritization only — qualified overflow stays visibly deferred.
Three separate dimensions, one explainable decision
northfield_v1 scores each account on three separate 0–100 dimensions: fit (ICP and firmographics), product usage (Northfield Incident Management signals), and buying intent (pricing, SSO, security, and admin activity). No opaque universal score collapses them.
Before any human action, visible gates decide the route — hand raiser, acquisition sales-assist, monitor, self-serve, manual review, or expansion — and assign the right owner with a real SLA. Weekly capacity is a separate decision: an account can qualify for a human action yet be deferred when the team is at capacity.
The naive comparator
combined_score_naive_v1 collapses the same three scores into one number — round(0.30·fit + 0.40·usage + 0.30·intent) — and routes any account scoring at least 60 into a single generic Sales queue. It has no hard blocks, no prospect/customer distinction, and no separate plays. It is a transparent educational comparator for showing why one number is often the wrong abstraction.
How they compare on the same 120 accounts
Both policies evaluate the identical synthetic account set at the same weekly capacity (default 12). The strip below is computed live from the current run.
- Qualified
- 25
- Routed
- 12
- Deferred
- 13
- Precision
- 80.0%
- Recall
- 35.7%
- False-positive rate
- 7.8%
- Routed yield
- 66.7%
- Qualified
- 22
- Routed
- 12
- Deferred
- 10
- Precision
- 86.4%
- Recall
- 33.9%
- False-positive rate
- 4.7%
- Routed yield
- 100.0%
The synthetic outcome comparison is illustrative, not a claim of real predictive superiority. The naive queue’s 100% routed yield at the default capacity of 12 is a small-sample artifact — all twelve routed accounts happen to carry a synthetic positive outcome. The gated policy is designed to provide context-specific plays, hard blocks, ownership, SLAs, rejected alternatives, and explainability — not to win a metric contest.
Limitations
Scores are expert-defined demonstration hypotheses, not validated predictors. All people, companies, usage, and outcomes are fictional or synthetic; precision, recall, and yield demonstrate method only and are never presented as real business efficacy.