How to forecast ongoing spare-parts demand

Start with equipment still in use

Build an active equipment register by model, serial number, installation date, location and usage. Remove confirmed retirements and distinguish equipment awaiting installation from equipment operating at customers. Mark unknown status explicitly rather than assuming every historical sale remains active.

Map each equipment version to compatible part revisions and the quantity fitted. Record component replacement dates because a replacement component can be younger than its host equipment. Keep UK demand separate from overseas demand unless the parts and service arrangements match. Use the batch and serial traceability guide to connect these records.

Estimate failures from comparable equipment

Group equipment by age, usage and operating conditions. For each component, divide confirmed failures by the corresponding exposure, such as equipment operating time or component operating hours. Include equipment that has not failed in that exposure. Keep the exposure unit and forecast period consistent.

For each group, multiply expected active component exposure over the forecast period by its observed failure rate, then allow for whether customers will repair, replace the whole product or obtain parts elsewhere. Add scheduled replacements separately. Investigate installation damage, repeat repair attempts and isolated defective batches before using them as normal failure demand.

Where records are sparse, use comparable equipment and factory reliability information as provisional assumptions. Show lower-demand and higher-demand scenarios and identify what remains unknown. Do not assume older equipment always fails more often or treat a factory reliability claim as observed customer experience.

Keep intermittent demand visible

Maintain a dated request history by compatible part and customer location, including periods without requests. Distinguish genuine absence of demand from missing records. Capture unfulfilled requests and cancellations caused by unavailable stock, since shipments alone can hide demand. The stockout forecasting guide explains that data problem.

Compare the equipment-based forecast with an intermittent-demand method that estimates request size and the gaps between requests separately, such as Croston's method. Use the result as an average demand estimate, not a promised repair schedule. Keep planned servicing, repairable-part returns and exceptional replacement campaigns separate so they are not counted again as ordinary failures.

Turn expected demand into a stock decision

Forecast across the period your replenishment decision must cover, including the ordering review interval and the route from factory release to usable UK stock. Confirm production availability, inspection, transport and receipt assumptions with the factory. The spare-parts planning guide supports the supply arrangements behind the forecast.

Choose stock cover separately from expected consumption. Consider downtime consequences, acceptable customer waiting time, compatible substitutes and replenishment uncertainty. A rarely requested part may warrant standby stock if its absence stops equipment. For repairable parts, account for repair turnaround and the chance that a returned part cannot be restored.

Deduct usable unallocated stock and credible incoming supply, while accounting for outstanding requests and reserved parts. Keep quarantined, incompatible and unserviceable parts out of available stock. Check factory pack quantities against the resulting requirement instead of increasing the demand forecast to justify a larger order.

Review results as the customer base changes

At each buying review, update installations, retirements, component ages, failure findings and factory availability. Replay previous forecasts using only information available at the time, then compare expected demand with actual requests over replenishment periods. Assess shortages, waiting time, emergency purchases and unused stock alongside forecast error.

Keep a forecast sheet showing the part revision, active equipment groups, exposure basis, observed failures, planned demand, uncertainty, available stock and proposed order. Record who approved the assumptions. As equipment leaves service or factory supply ends, review the remaining support population rather than carrying forward the old average. Use the last-time buying guide when supply is being withdrawn.