Churn alerts with reasons

A ranked list of specific, named risks per account — each with a severity and a sentence saying exactly what has or has not happened.

What it is

Rules over derived facts, not a black box. Examples: nobody signed in for 30+ days (high); usage dropped by half week over week (medium); no lead source connected after two weeks (high); website form connected but never fired (medium); email sending not set up (high); no estimate sent after 14 days live (high); estimates go out but customers never open them (medium); used to send estimates, none in 30 days (high); still a single-user account (low); three or more bugs in 30 days (medium). Each carries a detail line with the actual numbers.

Also called: at risk warnings · nobody has logged in · churn signals · why is this account at risk

See it
Churn alerts with reasons
CustomerBuildingAmountStatus
Ronan Petsch30×40 Garage$13,550Paid
Delia YoderBarndominium$2,530Draft
Marlin HooverRiding Arena$8,420Open
Gideon AltHobby Shop$15,830Open
Ronan PetschRiding Arena$5,000Approved
The Alerts card: severity-dotted rows like 'Website form connected but never fired — The source exists but no submission has arrived — likely not embedded.' SuccessPanel.tsx. Sample data — no customer information appears here.
How it works
  1. 1Every rule is guarded by how long the account has been live, so a brand-new account is not alerted for things it has not had time to do.
  2. 2Login-silence rules cascade so an account raises exactly one of the 7 / 14 / 30-day alerts.
  3. 3Alerts sort by severity, high first.
  4. 4The detail text carries real figures — '3 sent, 0 opened — check the send channel / spam'.
  5. 5Bugs filed by platform staff testing inside a tenant are excluded, because 'staff testing isn't the tenant's friction'.
Why we built it

The module's whole point is that a score alone is not actionable: it produces 'a list of ALERTS with the reason spelled out.' The staff-bug exclusion is commented directly, and the daily email that carries these says 'the alerts below say exactly what they haven't done.'

The problem
  • Health scores without reasons cannot be acted on.
  • Silent integration failures — a form connected but never embedded — went unnoticed for weeks.
  • Internal testing noise inflated a customer's friction metrics.
Sound familiar?
What you get
Every risk comes with the specific unfinished thing behind it.
New accounts are not alerted for things they have not had time to do.
The outreach conversation writes itself from the alert text.

See it on your own jobs

Twenty minutes, your numbers, no slide deck. We’ll build one of your real buildings in front of you and send you the estimate link at the end — yours to keep either way.

or keep browsing features →