Productized offer · Retention
Weeks before "we're moving on," the signs are already there — usage drops, invoices get paid late, replies go short, your champion quietly leaves. Churn Radar watches those signals across every account and flags who's drifting, with the evidence, while there's still time to do something about it.
The problem
Churn is almost never a surprise in hindsight. The signals were there — you just weren't looking at them, because they live in five different tools and nobody's job is to connect them.
The client doesn't complain — they just log in less, reply slower, pay later. Silence reads as "everything's fine" right up until it isn't.
Usage in your product, invoices in your accounting, sentiment in your inbox, org changes on LinkedIn. No single person sees the whole picture — so nobody does.
You "have a sense" of which clients are happy. That sense is calibrated on whoever you spoke to last, not on the accounts quietly drifting in the background.
How it works
Product usage, invoicing/payment timing, email & support sentiment, your CRM. Churn Radar builds one timeline per account from the tools you already use.
We look back at the clients you actually lost and read the signals they gave first. The model learns what "drifting" looks like for your business, not a generic SaaS benchmark.
A weekly health board plus real-time flags when an account crosses a threshold. Every flag links to the evidence and a suggested intervention. You decide; it watches.
The difference
Plenty of tools show a red/yellow/green dashboard. The difference is whether you can ask "why?" and get a real answer.
"Account health: 54/100 ⚠" — and nothing else. Why 54? Which way is it moving? What changed this week? What do I actually do?
— A number that makes you anxious without making you useful. You learn something's wrong at the same time you'd have noticed anyway.
"Northgate: drifting. Usage −40% over 3 wks, invoice paid 12 days late (was always early), champion Priya left 3 wks ago, new contact never onboarded. Suggested: re-onboarding call this week."
— Every flag is a story with receipts. You walk into the save-conversation knowing exactly what changed and when.
Pricing
No per-seat pricing. The fee is known before we start.
Questions
No — the score is the least interesting part. Every flag comes with the evidence underneath: which signals moved, when, and by how much, plus a suggested intervention. If you can't ask it "why?" and get a useful answer, it hasn't done its job. The whole point is walking into a save-conversation already knowing what changed.
Whatever you already have: product usage, invoicing/payment timing, email and support sentiment, CRM activity, and optionally public org signals like a champion changing jobs. In week one we connect the sources you use; if you don't have product-usage data (a pure services business), the model leans on communication and payment behaviour instead.
Yes — differently, but clearly. For retainers and project clients the signals are engagement-based: slower replies, fewer revisions requested, meetings declined, invoices paid later, a sponsor going quiet. Those are exactly the patterns Churn Radar learns from your own history of clients who left.
Precision is part of the agreed bar. A flag only fires when multiple independent signals move together, not on a single dip — one late invoice doesn't trigger anything, but a late invoice plus a usage drop plus a departed champion does. Week two is spent tuning that threshold on your history so false alarms stay rare enough that you keep trusting it.
Each one comes with a concrete suggestion — re-onboard the new contact, get a senior person on the next call, surface an unaddressed support issue. The system watches and advises; the relationship work stays human. Most clients find the suggested action is the part they'd have paid for alone.
Fit
Next step
Tell me how many clients you have and what tools the signals live in. I'll tell you honestly whether Churn Radar would have caught your last loss — and whether it'll catch the next one.
Replies within one business day.