Survey 1 · Batch 2 · 80 sequential cancellations
Where are we failing clients—and what can we fix?
The customer’s stated reason is evidence. It is not automatically the true reason.
We traced each case through the operating record, separated the cancellation trigger from the root cause, and asked a harder question: what did Simple.biz do wrong, and was there a realistic moment when we could have kept the client?
01 · Scope before conclusions
Eighty reviews. Seventy valid Website churn cases.
A cancellation list is not one clean population. We held out anything that would distort the Website churn diagnosis.
The denominator for all primary findings on this page.
Reported separately as acquisition or onboarding fallout.
Different service journey; excluded from Website conclusions.
Different service journey; excluded from Website conclusions.
02 · The true why
We know the root cause in 3 of every 10 cases.
The honest headline is not a single cause. It is a mix of proven company-service failures, customer and economic decisions, shared value gaps, and a large unknown bucket.
Root-cause certainty · 70 Website cases
211336
Supported 21 30%Possible 13 18.6%Unknown 36 51.4%
Inside the 21 supported roots
8Company-service4 quality/functionality · 3 PM follow-through · 1 service recovery
11Customer / economic4 business change · 4 affordability · 3 competitor or in-house switch
2Shared value gapLead performance where responsibility is shared or unresolved
What this means
We have a real, actionable quality and follow-through problem. We do not have evidence that company failure explains most cancellations. The unknown rate is itself a process failure in how we preserve account truth.
03 · What we did wrong
The supported failures concentrate in Development and PM.
These classifications describe the case record, not employee performance. Six cases contain supported failures in more than one department.
Process finding by function
Evidence status across 70 Website cases
“Possible” and “unknown” are shown separately so weak evidence never becomes blame—or accidental exoneration.
Sales0 supported
Development10 supported
PM / recovery10 supported
Billing1 supported
Supported failurePossible failureNo failure observedUnknown
Supported failures
Quality, functionality, completion, or launch-proof weaknesses recur in the record.
Supported failures
Follow-through, ownership, acknowledgment, or recovery handling is the other clear cluster.
Supported failures
Twenty-nine remain possible. Current evidence does not support a “Sales is clear” conclusion.
Supported failure
Billing is mostly an evidence gap, not a demonstrated churn cause in this cohort.
04 · Could we have kept them?
Four clear saves. Forty-nine plausible ones.
We reserve “could have retained” for cases where the record supports a realistic intervention. “Might” is a hypothesis for a save play—not a claimed win.
Retention opportunity · 70 Website cases
449107
Could have retained45.7%
Might have retained4970%
No credible action evident1014.3%
Unknown710%
Billing: cause or missing evidence?
49need a targeted FMS/QBO answer
vs1supported billing failure
Monthly billing belongs in FMS; QBO is supplemental for setup and related history. Pulling every transaction would add noise. The right question is whether a specific billing fact could change the diagnosis.
Early vs established11 cases canceled within 90 days; 59 were established. Three of the four supported early roots involved PM follow-through or website quality—but the early group is too small for a prevalence claim.
05 · Operating response
Six controls to reduce preventable churn.
Each action is tied to an observed failure or evidence gap and includes a leading indicator management can watch before the churn rate moves.
01Development
Evidence-backed launch
Require live-page QA, form delivery proof, customer revision closure, and explicit acceptance before a site is treated as complete.
Evidence basis10 supported Development failures · 4 supported quality roots
Leading indicatorUnresolved launch defects at delivery
02PM
One owner, one recovery clock
Route every stop, hold, dissatisfaction, value, or provider-switch signal to one accountable owner within one business day.
Evidence basis10 supported PM failures · 4 PM/recovery roots
Leading indicatorRisk signals without owner response in 1 day
03PM + SEO
Reconcile value at 30 / 60 / 90
Review traffic alongside qualified calls, forms, lead quality, customer follow-up, and commercial outcome—not traffic alone.
Evidence basis2 supported shared value/lead roots
Leading indicatorAccounts missing a signed value review
04PM + Accounting
Capture the cancellation truth
Record the stated reason, immediate trigger, true root cause, internal failure, save attempt, and evidence still missing.
Evidence basis36 of 70 root causes remain unknown
Leading indicatorCancellations closed without a causal review
05Accounting
Target billing evidence
Pull FMS recurring history and QBO setup detail only when the result could change the diagnosis; link the answer back to the case.
Evidence basis49 need a targeted lookup · only 1 supported billing failure
Leading indicatorDiagnosis-changing billing questions unresolved
06Sales → PM
One accepted promise
Preserve goal, scope, platform, timing, setup price, monthly price, and success language through the handoff and launch.
Evidence basisNo supported Sales failure · 29 possible · evidence remains weak
Leading indicatorDelivered files missing the accepted promise
06 · What the evidence permits
Act on what is supported. Investigate what is not.
This is a sequential cancellation cohort, not a randomized study and not a churn-rate denominator.
- Development and PM contain the clearest supported internal failures.
- Quality/functionality and PM follow-through are recurring supported company-service roots.
- Only four cases support a confident “we could have saved this client” judgment.
- Billing is currently more often a targeted evidence gap than a supported cause.
- That company failure caused most cancellations.
- That Sales is cleared because no supported Sales failures were found.
- That 49 “might” cases would convert into 49 retained clients.
- Which failures distinguish churned customers from retained customers without an active-customer comparison set.