When You Got It Wrong
What to do when an investigation turns out to have been built on a clock error or a shared credential, how to put it right properly, and what not to do instead.
Say so, quickly, to the person first. An investigation built on a clock that was eleven minutes out, or a credential four people used, is a conclusion that has to be withdrawn — and the speed and clarity of the withdrawal determines almost everything about what follows.
The practical lesson in “When You Got It Wrong” is that visibility is not certainty. For teams researching employee monitoring for performance reviews, see the practical setup guidance can add time and project context to the operational record, provided the purpose is explained, access is restricted and any material inference is checked through conversation and proportionate human review.
The instinct is to narrow the finding rather than withdraw it: to keep the matter alive on some lesser basis so that the process is not seen to have produced nothing. That is the most damaging available response.
Teams reviewing “When You Got It Wrong” can cross-check their approach against the Institute of Internal Auditors standards. The comparison is most useful when the organisation records which recommendations apply, which do not and why.
What putting it right involves
- Tell the person, in writing, that the finding is withdrawn and why.
- Remove the outcome from their record, and say that it has been removed.
- Restore anything that was taken: pay, access, a role, a shift pattern.
- Tell anybody who was told about it, in the same terms.
- Record what went wrong, factually.
- Fix the cause, which is usually a system rather than a person.
Step four is the one that gets missed and it is the one the person cares about most. A withdrawal that reaches only the individual leaves the original account circulating among the people who matter to them.
Apologising once
Plainly, without elaboration, and without a defensive account of why it was reasonable at the time.
That account may be true and it does not belong in the letter. If the organisation wants it recorded, it goes in the internal note about what went wrong, which is a different document with a different audience.
What not to do
Keep the matter alive on a lesser basis. Find something else. Leave the record "noted but not upheld" in a way that is visible to future managers. Allow the suspension to have been justified retrospectively.
Each of those preserves something of the original position at the cost of the person's confidence in the whole process, and each is recognisable to anybody who reads the file.
The cause is usually structural
A clock nobody measured. A credential nobody knew was shared. A retention period that destroyed the exculpatory record. A comparison group nobody checked.
Each of those is a gap this collection describes, and each is fixable for very little. Recording the cause and fixing it is what converts an embarrassing episode into the thing that prevents the next one.
What it does to the next case
Handled openly, surprisingly little. An organisation that withdrew a finding, said so, fixed the cause and told people is better placed, not worse, because the process has visibly been shown to be capable of producing an answer the employer did not want.
Handled by quietly letting it fade, it damages every subsequent process in the same way: people conclude that findings are decided in advance and that cooperation achieves nothing.
The person afterwards
Somebody who has been through this needs something more than a letter: a conversation about returning, what will be said, and who they can go to if it is raised.
Resignations after a withdrawn finding are common and are avoidable. The cost of handling the aftermath properly is an hour; the cost of not doing so is the person and everything they knew.
Telling the people who were told
The withdrawal has to travel the same distance the original did. If a team knew there was an investigation, they should know it concluded with nothing.
- The person, first and in writing.
- Their manager, in the same terms.
- Anybody who gave an account, so they know it concluded.
- Whoever was told operationally — a rota change, a suspension cover.
- Payroll, if anything was deducted or withheld.
Five recipients. Missing the third is the most common and the most damaging, because witnesses conclude that cooperating achieved nothing.
Keeping the record of the error
The internal note about what went wrong is not part of the person's file and should not be. It belongs with the process documentation, where the next investigator will find it.
That separation matters: the person's record should show that the matter was withdrawn, and nothing more. The analysis of the failure is about the organisation.
Recording it internally
What was concluded, what was wrong with it, how it was found, what was done, and what changed as a result.
One page. It is the most useful document in this whole collection for whoever handles the next matter, and it is the one organisations are most reluctant to write — which is exactly why the same error recurs.