Control · C73

Define initiative stop rules

Question addressed

Which evidence justifies continuing, correcting, reducing or stopping an initiative?

Direct answer

The grid combines economic threshold, maximum learning time, data quality and value of the remaining option.

Decision supported

01Continue, correct within a deadline, reduce exposure or stop.

02The rule, the evidence and the next review date stay visible.

How to use this

Move from the question to the deliverable.

01Starting point

Which evidence justifies continuing, correcting, reducing or stopping an initiative?

02Decision prepared

Continue, correct within a deadline, reduce exposure or stop.

03Deliverable to use

Stop-rule grid

Decision grid · SQ-73

Decide to stop before sunk costs take over

The stop rule is written before the main investment. It combines a minimum result, a learning deadline, the quality of the evidence and the value of the remaining option.

Working context

Grid applied to four initiatives after six weeks of execution and €185k committed.

  1. 01Pre-commit

    Fix thresholds, deadlines and exceptions before launch.

  2. 02Observe

    Compare the evidence obtained with the minimum result and the expected deadline.

  3. 03Value the option

    Estimate what can still be learned or recovered through a limited tranche.

  4. 04Decide

    Continue, correct within a deadline, reduce or stop without renegotiating the criteria.

Original element

Four initiatives against their stop rules

A red threshold overrides the overall score. An exception requires new evidence and a capped budget.

InitiativeExpected thresholdObservedEvidenceRemaining optionDecision
ABM pilot8 activated accounts11StrongProgressive scaleContinue
Webinars18% attendee→MQL9%ReliableLowStop
OnboardingActivation ≥ 45%39%Partial€15k UX testCorrect within 3 weeks
Sponsorship120 qualified leads74Incomplete trackingAudience recoveryReduce
Decision interpretation

01The webinars are stopped despite €48k already committed: the evidence is reliable and the remaining option is weak.

02Onboarding receives one further tranche of €15k with a three-week deadline.

03Sponsorship is reduced, because measurement uncertainty forbids a conclusion but does not justify the initial exposure.

Fictitious case. A stop rule must distinguish an absence of effect from an inability to measure the effect.

Coherence controls

Minimum conditions before validation.

01Constant scope

The status, the evidence and the owner must be available for review.

02Documented cause

The status, the evidence and the owner must be available for review.

03Assigned decision

The status, the evidence and the owner must be available for review.

04History kept

The status, the evidence and the owner must be available for review.

Usable deliverable

Output produced

Stop-rule grid

Explore the decision further with an interactive control, or start from a structured template to produce your deliverable.