Documenting a redesign that did not improve the outcome

Record an unsuccessful redesign so the team can see what changed, what was expected, what happened and what remains unknown. Keep the observed outcome separate from an explanation for it. A result that missed the goal may call for a correction, another test or a different approach; the record should make that decision possible.

State the decision and expected result

Start with the customer task and why it changed. In a hypothetical appointment service, a revised confirmation might be intended to help people know whether a rescheduled booking is final. Record the original uncertainty, the version released, the audience exposed and the success condition agreed before release. Do not rewrite the original aim to match the outcome.

Keep the chronology brief: when the baseline was gathered, when the redesign became available, what other material service changes occurred, and when outcomes were checked. Record who approved the change and who owns the review.

Show the result with its limits

Describe the measure and its denominator, then show what was observed. Separate completed, failed, still-open and unknown outcomes. Put feedback beside the service result; do not substitute one for the other. If people say the new confirmation is clearer but still receive an unsuitable appointment, the clarity result cannot stand in for the booking outcome.

Check whether the periods or groups are comparable. A changed invitation point, case mix, staff procedure or booking rule may affect the result. Record such differences beside the conclusion. If the evidence is too thin to determine whether the outcome improved, say so rather than labelling the redesign a failure or success.

Examine possible explanations

Use a compact investigation record:

For the hypothetical appointment service, the message might be unclear, the booking rule might reject some requests, or the confirmation might be sent before a decision is final. These are distinct hypotheses. Do not present the most convenient one as a finding until relevant evidence supports it.

Hypotheses for the Redesign's Limited Success

Make the next action explicit

If customers have received a wrong or misleading result, route those cases for appropriate help while the wider cause is investigated. Decide whether to correct the redesign, limit its use, restore an earlier route where feasible, or gather evidence before another change. Record who can authorise the action and how affected customers will be updated.

Close the review with a checkable learning statement: which assumption was unsupported, which finding is established, and what the next version must demonstrate. Preserve the original measure and contrary cases. That history lets the team test a revised idea without repeating the same mistake or treating delivery as improvement.

Next Steps Following the Redesign Review