Our approach

The rubric behind every score, alert, and QBR brief.

Product adoption
Engagement
Support
Sentiment
Commercial
82example score
At risk<60
Watchlist60–79
Healthy80+

5 inputs, 1 weighted formula, 1 score, 1 band — alerts fire on watchlist / at-risk only.

Five signals, not one gut check

Product adoption

Are they actually using it?

Depth & breadth of usage

Engagement

Is the relationship still alive?

Contact cadence & stakeholder breadth

Support

How much friction are they hitting?

Ticket volume, severity, resolution

Sentiment

How do they feel about it?

Tone from calls, convos, NPS

Commercial

Where is the account heading?

CRM stage, billing, renewal posture

One score, three bands — fixed forever

Fixed platform-wide — never tunable per workspace.

06080100

At risk

Below 60

Watchlist

60–79

Healthy

80+

Metrics older than 60 days decay to half weight — stale data isn't healthy, it's unmeasured.

What's fixed, what's configurable

The definition of "healthy" never moves — which signals earn it can.

Fixed for every workspace

  • Health score bands (80 / 60 / below)
  • The five metric categories
  • The five alert types
  • Expansion is CRM-signaled, never inferred

Configurable per workspace

  • Weight given to each signal
  • Renewal-window lookahead
  • No-contact threshold
  • Health-drop sensitivity

Alerts are a triage rubric, not a notification feed

5 fixed alert types, severity-sorted — not chronological.

Urgent support ticket

Critical— Open P1 — skips the queue

Health score dropped

High

Sharp drop, same-week check-in

Renewal window

High / Medium

Unsigned, inside the window

Champion or exec departed

High

Best-documented churn predictor

No contact

Medium— At-risk account, gone quiet

Expansion is a different question than "healthy"

A high score means not-at-risk, not ready-to-expand — expansion is set only from an explicit CRM signal, never inferred from health.