Post-Kaizen Sustainment Audit

Why it matters

On the last day of a kaizen event, a new layout may have been set up, a document prepared and training scheduled. These are necessary deliverables; they do not yet show that the change is being kept under working conditions. When the event team's support is withdrawn, material presentation, tool access or task handover may change. The sustainment audit investigates this change. The aim is not to blame the worker for going back to the old method but to understand under which conditions the new method can survive. The audit should start by observing the critical steps that must be done. Then the decision should be made together with the output and quality results measured on the same scope. That way a document merely hanging on the wall or an activity list being closed is not presented as success; the real use of the method and its effect become visible.

The conformity rate shows in what share of the audited applications the defined method was kept. It is not the same measure as product quality. The result can deteriorate while the method is applied correctly; for example, input conditions may have changed or the first solution may have addressed only part of the problem. Conversely, in a period when the method is departed from, the output can look good; extra support or a temporarily easy product mix can mask the result. For this reason application evidence and result evidence are held side by side. Using one in place of the other creates false confidence. Comparing periods without knowing the denominator of the rate and the scope of the audit is also unsafe. Each check should include the same critical steps, similar product conditions and a stated shift scope.

The 30, 60 and 90-day checkpoints are not automatic success dates. These dates are opportunities planned to examine how the change behaves when support has decreased and ordinary working conditions have returned. A single good day does not prove long-term stability. Besides the value at the time of the check, the direction relative to earlier values matters. If conformity stays the same while the good-product output rate falls and the nonconformity rate rises, lasting success cannot be claimed by looking at the percentage alone. Causality is not settled by these three observations either; changes in product, equipment and replenishment conditions are investigated. The task of the audit is to make the need for investigation and correction visible, not to produce a certain cause from limited data.

This lesson audits the preservation of the change after kaizen. The detailed management routine of the daily results meeting belongs to M28; we do not design that whole routine here. The check owner, the data source, the acceptance condition and the path to follow in case of deviation are defined. The audit result can lead to a decision of acceptance, conditional monitoring or rework. If a safety or critical quality condition has been broken, the application is not accepted even if the overall conformity percentage is high. The limit of the success claim is written clearly. A result found acceptable for the area examined is not automatically generalized to other products, shifts or lines that look similar. The sustainment decision is limited to the scope the evidence represents and can be reopened at a new deviation.

While preparing the sustainment audit, how the old measurement was made is also checked. If all products were counted in the baseline data but only good products are counted in the follow-up, the difference in rate does not represent the same concept. If the scope of the observed time has changed, the comparison is limited even if the rate can be calculated. Instead of filling in a missing definition later with an estimate, a new and consistent reference period may be needed. This decision is documented openly. Expecting a positive result about the change does not make the old data more reliable than it is. For this reason it is important for the follow-up table to include the measurement definition, the record source and a scope note. That someone else can reach the same result from the same numbers shows that the audit was not prepared only for presentation. Sustainment evidence also requires traceability of the data; a nice percentage whose source cannot be explained is not a sound basis for a decision.

Formula