Job renewal-risk-list-with-evidence-for-each-account · revised 12 October 2026
See which accounts renewing soon look at risk, with evidence for each
A list of accounts renewing in the window you set (for example 90 days) that show any agreed risk signal, each with the evidence and its date. Read-only; the account owner decides.
You might be seeing
- An account cancels and nobody saw it coming
- Signals sit in four systems and nobody looks at them together
No passwords, keys, card details or admin invites needed to start.
What usually happened
Churn signals such as no logins, more tickets, an overdue invoice or a leaver sit in different places. Looking at them for every renewal by hand does not happen, and a bare score nobody can explain gets ignored. A list that is noisy or falsely calm costs trust.
Who it’s for: Customer-success lead at a B2B firm with annual contracts.
Usually starts when: A surprise cancellation at renewal makes everyone ask what the signs were.
The result: You receive a list of accounts renewing in the window you set (for example 90 days) that show any agreed risk signal, each with the evidence and its date. It is read-only, and the account owner decides what to do.
Check whether this job fits
Answer these without sending files or logins. Nothing is submitted unless you choose to contact us.
What you get
- The rule sheet
- The risk list with evidence
- The seeded test results
- The back-test report
Included
- One CRM and the signal sources you name, up to four
- A rule sheet you approve: signals and thresholds
- A list of renewing accounts with any signal, each with evidence and date
- A seeded sandbox check, one hit and one near-miss per signal
- A 12-month back-test reported with sample sizes and no promise
Not included
- A churn probability or a forecast
- Contacting any customer
- Changing any record
- Extracting renewal dates from contracts: that is a separate job
- A promise that an empty list means no risk
How we know it’s done
Agreed with you before work starts. Each check produces evidence you keep.
In seeded sandbox accounts, each signal's hit and near-miss are classified correctly: 100% of the seeded accounts.
Evidence: The seeded accounts and the output.
In a sample of 30 live rows, 0 evidence items mismatch the source system.
Evidence: The sample with your marks.
A 12-month back-test reports how many churned accounts would have been listed 60 or more days ahead and how many listed accounts did not churn, with sample sizes and no promise.
Evidence: The back-test table.
Listed plus not-listed accounts equal all renewing accounts in the window.
Evidence: The count reconciliation.
Sign-off. You re-run the test yourself, check the list of severe errors and sign off in writing. Payment follows sign-off.
If it fails. If the agreed pass mark is not met, we say so, hand over the results and the failures, and do not call the output working. You do not pay for this scope; we agree whether to stop or re-quote.
When it fits, and when we stop
It fits when
- Renewal dates are held in the CRM; if they are only in contracts, extract them first
- At least 50 accounts, so the back-test has some cases to count; with so few it is a description, not a measure
- Signal history can be exported as at past dates, so the 12-month back-test can be run
- A named customer-success owner approves the rules
- Real records are read only after a written agreement that covers how your records are processed and a separate access step you control; until then we test on invented or redacted data.
We stop and tell you if
- Fewer than about 50 accounts
- You want a churn probability or a forecast
- The signals are not available to us in any export
- Signal history cannot be exported as at past dates
What could go wrong
The job reads exports and writes a list, so there is nothing to undo.
Scroll the table sideways to read it all.
| Risk | How we handle it |
|---|---|
| Noise or false comfort | The list says it is not a forecast and reports the back-test with its counts: listed accounts that did not churn, and churned ones that were not listed. |
| Signals are read as proof of intent | Each row shows the evidence and its date only. |
| Usage data is missing | We use only the signals you can export and say which are absent. |
A separate automated reviewer checks the list, the evidence and the back-test counts before hand-over, and no person on our side reviews unless your agreement names one. Your customer-success owner approves the rules and decides every action. At launch the work is largely automated, and we say so.
How we deliver
We arrange the work and independent review, then show you the result against the agreed checks. You keep authority over your systems.
- Agree the signals, the thresholds and the reporting window in writing
- Receive the exports by the agreed route
- Seed sandbox accounts, one hit and one near-miss per signal, and run them
- Run the rules on the real exports and write the evidence per account
- Back-test on the last 12 months and report the counts
- Independent automated review, then hand over the list
This is a one-off job, not a subscription or emergency cover. We confirm eligibility, price, a start window and a delivery date before you accept, and work starts only after agreed inputs and access are in place. Nothing is sent, merged, posted or changed in your live systems by us, and an enquiry creates no charge or booking.
Need to keep it working?
Discuss a weekly renewal-risk list that keeps running.
Ongoing work is separately scoped and quoted: no monitoring, response-time guarantee or automatic subscription is included in this job.
Mention the responsibility you need in your enquiry.
What you can check
This is a new service. We have not delivered this job for a client yet.
Other ways to get this done
- Your customer-success tool may already flag low usage or overdue invoices. Check what it shows and whether it gives the evidence.
Questions
Will it predict churn?
No. It lists accounts with agreed signals and shows the evidence. It does not give a probability.
What if the list is empty?
That is not a promise of safety. It only means no agreed signal fired.
Does it contact customers?
Never. It only reads the exports you give it and returns a list.
Send an enquiry
Send us
- The signals you would use and where each one lives
- Roughly how many accounts renew in a year
- How renewals are recorded today
Later, once you agree
- Exports for accounts, renewals and each signal source, or read-only users you can revoke
- The approved rule sheet
- A secure handoff route agreed before any record is shared
You keep every system. We read agreed exports and return the list. Which service reads your records is agreed in writing before any real record is shared; the entry test uses invented or redacted data, or your own provider account.
Email fallback: open your mail app
If website submission is unavailable, review and send the fallback email yourself. An email fallback is not a website receipt. Or write to hello@syntheticindustry.ai with “renewal-risk-list-with-evidence-for-each-account” as the subject.