A shorter average lead time is an attractive finding in a periodic review. Before using it, I would ask which orders entered the calculation and how long the orders still open have been waiting.

If the average covers only records that reached a particular milestone, faster orders may appear first while slower ones remain outside it. The figure can be useful without describing the performance of the entire order population.

Give the average a definition

State the start, end, included scope and cut-off. Starting at requirement confirmation is different from starting at formal order receipt. Estimated departure, actual departure and customer receipt are also different endpoints.

If the source contains estimated departure dates, describe the result as a scheduling measure. Do not turn those dates into evidence of actual departure. Before comparing periods, check whether samples are excluded consistently and whether the product mix is comparable.

Keep open orders beside the average

What to examineQuestion to ask
Records reaching the specified milestoneWhich start and endpoint, supported by what evidence?
Records still openHow long have they waited, and is the original commitment overdue?
Remaining planWhat is the expected milestone, basis and responsible owner?
Comparison scopeAre product mix, exclusions and cut-off consistent?

An unfinished order should neither be counted as finished nor disappear because it cannot enter a completed-cycle average. Show its elapsed waiting time and risk separately. Its estimated total cycle is not an actual completed cycle.

Explain both the observation and its limits

The records that reached the specified milestone show a shorter cycle. Open orders are listed separately with waiting time and expected arrangements. Differences in product mix and state remain for the next review.

This preserves the useful finding while keeping the remaining work visible. Two averages alone do not establish that one inventory measure caused all the improvement. Scheduling, product mix, order timing and outstanding backlog also need examination.

For training, use entirely fictional orders in which short cycles finish first while long cycles remain open. Ask a colleague to calculate the completed average, then explain why it cannot support the statement that every order is now faster.

Related: Replace “in progress” with order-level variance.


Drawn from questions raised in a periodic work review. Identities, business figures and project details have been removed. Checklists and sample wording are for similar situations; they do not establish a completed outcome.