How-To·5 min read·Published 29 September 2026

    How to measure stockouts and excess inventory consistently

    Quick answer

    Define a stockout before counting it, use the same definition in both periods and document the time window. Measure excess inventory separately, using an agreed rule and cost basis. Keep estimates, counts and percentages clearly distinguishable.

    A before-and-after result is only useful if everyone is counting the same thing. One team may count a late order as a stockout; another may count each missing order line. Both numbers can be valid, but they cannot be compared without explaining the difference.

    Set the definitions with the buyer and operations team before changing the planning process. This gives the business a useful baseline and prevents a later improvement from being overstated or missed.

    Choose the event you will count

    For committed-order stockouts, define when an order is considered promised and what constitutes insufficient stock. Decide whether the count is by order, order line, SKU or job. Record the required date and cause so a missing item is distinguishable from a courier delay.

    For a service team, the useful event may be a job delayed by a short kit. For an online store, it may be a committed order that cannot be fulfilled. Use the measure that describes the actual operating problem.

    Use comparable windows

    Record the start and end of the baseline and comparison periods, the go-live date and any phased rollout. Compare equal periods where practical, and note changes in order volume, product range, seasonality and unusual wholesale commitments.

    A snapshot taken at diagnosis is a different measurement from the number of incidents over the next quarter. Label it as a snapshot when that is what the source shows. Do not silently turn it into a 90-day baseline.

    A worked calculation

    Illustrative example: if the same definition produces 20 stockout events in the baseline period and 8 in a comparable later period, the reduction is (20 minus 8) divided by 20: 60%.

    If order volume changed substantially, show the volume beside the count. You may also report a rate, such as affected orders divided by total committed orders, provided both numerator and denominator use consistent definitions. A zero baseline does not support this percentage-reduction calculation.

    Measure excess stock separately

    Agree what excess means for this business. It could be stock above an agreed coverage threshold, obsolete items or quantities beyond expected requirements after existing commitments are considered. Apply the same rule in both periods.

    Value the relevant units at an agreed cost basis rather than switching between cost and sales price. Explain whether damaged, obsolete or allocated items are included. Stock locked against future orders is a specific allocation measure; it should not automatically be relabelled as every form of excess inventory.

    Keep the source record

    • Metric name and exact definition.
    • Baseline and comparison dates.
    • Source system, extract date and responsible person.
    • Count or value, units and any denominator used.
    • Changes in demand, product range or rollout scope.
    • Whether the result is measured, calculated or estimated.

    A reported time saving may be an estimate from the team. That can still be useful, provided it is labelled and the method is explained. It is not the same evidence as a timestamped operational record.

    Read the client stories in context

    The RMA result is labelled diagnosis versus since go-live. The Car Park King short-kit result is expressed per quarter. Whatsafety's manual-admin reduction is an estimate. Their different measurement bases should remain visible.

    Use this checklist alongside the implementation data checklist. The aim is a better buying process and evidence your team can explain.

    Frequently Asked Questions

    Should we count stockouts by order or by SKU?

    Choose the unit that represents the business problem, document it and use it consistently. Order counts and SKU counts answer different questions and should not be combined.

    Can estimated admin savings be included in a case study?

    Yes, when they are clearly labelled as estimates and separated from directly measured counts, values and delivery dates.

    Make your next buying decision clearer.

    See where stockouts and excess inventory are costing your business. Start with a free stock diagnosis.

    Get a free stock diagnosis

    Related articles