How to fix manager mistakes: 90-Day Competency Rule vs. Delayed Review

From that table, the only concrete figure is Gallup's report that global engagement fell for the second straight year to 20%. The Columbia Business School executive leadership programs are listed, but the supplied grounding shows no manager-mistake outcome from them. The ACE Verification glossary defines verification leadership without a 90-day rule. None of these items converge on a manager-mistake fix number.

With no grounded convergence, the audit rule is strict: do not borrow the 20% engagement figure to justify any gate length for a manager mistake. For any candidate manager mistake, you must ask for source-backed evidence of its outcome type before choosing a gate. If the only available evidence is a general trend, it cannot satisfy that requirement on its own.

The final check is this: if you cannot point to a named source from the supplied grounding that gives a manager-mistake outcome for that exact mistake type, you must not assign a fix number. Treat that absence as a demand for observable behavior evidence, not a reason to guess. No grounded source converges on a manager-mistake fix number, so the gate decision must be earned by evidence, not by trend figures.

Compare gates by error and review cost

This section alone presents the options comparison table and names the winner. For the purpose of choosing between gates, the comparison must be made against the exact conditions that can be met, not against preference.

With that comparison in mind, the 90-Day Competency Rule is the winner when all three checkpoints are observable, and Delayed Review is the winner only when the outcome lag exceeds day 90. Any other split between them is a gate error, not a trade-off of convenience.

For either gate, the cost formula is review hours × loaded manager hourly rate + repeat-error exposure. Treat this as a method, not a universal quote: pull loaded manager hourly rate from your own local payroll inputs, count only review hours tied to this manager-mistake, and account for repeat-error exposure against the work that would be pulled to correct it again. Do not borrow foreign benchmarks to fill any missing piece. If any of those inputs cannot be pulled from your local payroll inputs at the time of choice, you cannot compute a full dollar figure and must state the missing input before picking a gate.

To compute the cost of being wrong, run the same formula twice against the choice you are about to make: once for the gate you would pick, and once for the gate you would have picked if the condition flipped. The difference is the avoidable cost of the wrong gate. The wrong gate does not just add review hours, it can create repeat-error exposure that would not exist under the correct gate, or it can create unnecessary review hours that would not exist under the correct gate. Because of that, the decision must be driven by the condition above, not by trying to minimize review hours alone.

Limit the rule and test a worked case

This section alone provides the edge-case table and the copy-usable Manager Mistake Evidence Log. Before assigning any manager mistake to a gate, sort it by observability. If the fix is a reversible behavior that can be pulled from weekly samples, the 90-Day Competency Rule still wins. If the true proof is a lagging outcome beyond day 90, Delayed Review is the only one of the two gates that can hold that proof. If the mistake is an irreversible safety or compliance mistake, neither gate alone is enough, and you must require supervised sign-off.

To force that discipline, use the Manager Mistake Evidence Log for every pilot entry, and fill it with real pilot entries, not placeholder evidence. The columns must be: mistake ID, behavior definition, baseline sample, day-30 clean cycle, day-60 clean cycle, day-90 clean cycle, outcome owner, decision. For each row, the decision must be traceable to the samples, and if Delayed Review is chosen, the outcome owner must be the same person named for that lagging outcome.

Run three checkpoints on purpose. The baseline sample must be date-stamped before any coaching action counted toward the fix. Each clean cycle must be a scored sample pulled from the exact behavior definition, not a general impression. If you cannot complete all three clean cycles by day 90 with date stamps, do not stretch the 90-Day Competency Rule to cover a lagging outcome. Only move to Delayed Review if you can name that lagging outcome and pre-register its outcome owner at the moment of that switch.

For irreversible safety or compliance mistakes, block any gate choice until supervised sign-off is assigned. Do not treat "no failure seen" as proof for a lagging outcome. If any column of the log cannot be filled with a real sample from your pilot, the edge case must be re-sorted to Delayed Review or to supervised sign-off. That is the only way to limit the rule to cases where three checkpoints are actually scorable inside 90 days.

Apply five if-then decision rules

Rule three: if the mistake is irreversible or safety-critical, then require supervised sign-off plus the gate. The gate itself is not removed by this, but the removal of supervision cannot happen without that sign-off. Rules one and two are the ordering steps at the end of this section: define the exact behavior to score, and confirm it is visible in weekly samples inside 90 days, before any gate is assigned. Rule four: if the manager mistake type is undefined, then do not assign either gate yet. Without a defined behavior, there is nothing to count for a pass, and there is nothing to verify for a post-day-90 check.

Rule five: if you cannot pre-register a specific lagging outcome and its owner for a post-day-90 check, then do not switch to Delayed Review. The switch must be justified by that named lag, not by workload or preference. If the only reason to push past day 90 is that it is more convenient, the 90-Day Competency Rule cannot be bypassed. The owner must be known at the moment of choice, with the review date tied directly to that outcome.

Apply these rules in strict order for every manager mistake: define the behavior, test weekly visibility against day 90, count toward three date-stamped clean cycles, check for irreversibility or safety-critical status, and only then choose the gate. If any rule fails at its step, the default of the 90-Day Competency Rule is not met. This keeps the decision to observable evidence and prevents stretching the calendar gate past the proof it requires.

Frequently Asked Questions

Can the 20% engagement figure be used to justify a 90 days gate for a specific manager mistake?

No, the audit rule is strict: do not borrow the 20% engagement figure to justify any gate length for a manager mistake.

Before choosing between gates, what must be obtained for any candidate manager mistake?

For any candidate manager mistake, you must ask for source-backed evidence of its outcome type before choosing a gate.

Do the Columbia Business School executive leadership programs provide a grounded manager-mistake outcome to support either gate?

The supplied grounding shows no manager-mistake outcome from the Columbia Business School executive leadership programs.

Does the ACE Verification glossary provide any support for a 90 days rule?

The ACE Verification glossary defines verification leadership without a 90-day rule.

If the only evidence available is a general trend, can it satisfy the requirement on its own?

If the only available evidence is a general trend, it cannot satisfy that requirement on its own.

What must happen if there is no named source from the supplied grounding that gives a manager-mistake outcome for that exact mistake type?

If you cannot point to a named source from the supplied grounding that gives a manager-mistake outcome for that exact mistake type, you must not assign a fix number and must treat that absence as a demand for observable behavior evidence, not a reason to guess.

Quick answers

Also worth reading: LMS Competency Mapping: 3 Platforms, xAPI Depth vs SCORM Fallback 2026: LMS Competency Mapping: 3 Platforms, · 2026 Peer Coaching Academy: Architecture, Mistakes & Tactics: 2026 Peer Coaching Academy: Architecture, · How to Use Assessment Data to Close Leadership Skill Gaps: How to Use Assessment Data