Diagnose and Reverse a Falling Health Score
When an account's health score falls, FunnelStory already has the "why" — the factors behind the score, the open concerns, and a way to test what would move it. This routine takes you from noticing the drop to a concrete intervention in a few minutes.
1. Spot the drop
Health shows as a score from 0–100 with a tier chip — Critical (45 or below), At risk (46–54), or On track (55 or above) — on the Accounts view and in Focus Areas. A move into Critical or At risk is the trigger for this routine. See the health score badge.
2. Read the driving factors
Open the account and go to Renewal → Driving Factors. It lists what's pulling the score down and what's holding it up, each a concrete factor (an overdue renewal, a weak relationship, falling usage, open risks). This is your diagnosis — see Account predictions and the Predictions overview.
3. Check what's open
Confirm the story against the account itself:
- Needle Movers — the open risks (pricing, competitor, champion leaving) driving the drop, with their source excerpts.
- Account Dossier → Active Concerns — what's working against the relationship in the last 90 days.
- Adoption → Tickets and the last meeting — open escalations and the most recent conversation.
4. Test an intervention with What-If
Use What-If analysis to see which change would move the score most: adjust a factor (for example, add an engaged contact, or lift a usage measure) and watch the score update against the distribution of similar accounts. Pick the one or two factors with the most lift — those are where your effort pays off.
| If the biggest factor is… | A good intervention |
|---|---|
| A weak or single-threaded relationship | Bring in a second stakeholder; see Relationship maps |
| Falling product usage | A targeted enablement or re-onboarding touch |
| An open pricing or competitor concern | Address it directly before renewal |
| An overdue renewal | Get the renewal conversation scheduled |
5. Act
Turn the diagnosis into a next step: create an Action Card or task with the intervention and a drafted email, or let a needle mover's Investigation and Triage chat draft it for you. Record the plan as a note so the next person sees it.
Mind data confidence
A high prediction confidence with low data confidence means the model found a pattern on thin data — treat those with appropriate skepticism and lean on what you can see directly. See Confidence ratings.
Related
- Account predictions — the health badge and driving factors.
- What-If analysis — test a change before you make it.
- Prepare for a Customer Call — the pre-call routine for the conversation that follows.