Launching an aquarium light in a new country creates an immediate service question: which replacement parts should be available locally, and how many? There is no field history for that exact market, so copying one percentage across adapters, mounts, controllers, cables, and optical covers seems convenient. It is also misleading. Demand depends on the installed population, part criticality, replenishment time, compatibility, repair authority, storage behaviour, and the reasons customers actually need support.
Start with the planned installed base over time. A launch of 500 units followed by rapid growth creates a different service curve from one small specialist shipment. Record sell-in and, where reasonably available, sell-through or activation timing rather than assuming every carton enters service immediately. Separate models, lengths, regional power configurations, and revisions. A spare that fits the first shipment may not fit a later change even when the retail name remains familiar.
Classify parts by service impact and who may replace them. An external mounting foot or approved adapter may be designed for straightforward replacement. A sealed optical or electrical assembly may require authorised service or complete-unit exchange. Availability should never imply that customers may open protected or mains-powered equipment. The product design, instructions, warranty policy, tools, test process, and local competence must support the stated repair route.
Lead time can be more important than unit price. A low-cost bracket produced only with quarterly factory orders may justify a larger local buffer than an expensive module that is rarely used and can be shipped rapidly. Include manufacturing lead time, international transport, customs, warehouse handling, order cycle, and desired service response. Consider minimum order quantities and pack multiples. Document the assumptions so the next planner understands why the starting stock is not simply a percentage of sales.
Compatibility deserves its own control. Give every spare a code and define the models and revisions it serves. Photograph connectors or mounts only as supporting information; use drawings, bills of materials, and approved compatibility records for decisions. If a controller or power supply changes, decide whether old and new parts can be substituted, require a kit, or must remain separate. Obsolete stock grows quickly when the forecast ignores engineering change.
Storage life and conditions vary. Adhesive pads, batteries, gaskets, labels, and some packaging materials may need rotation or environmental limits. Metal parts can corrode in poor storage, and clear covers can scratch when stacked loosely. Record any manufacturer-defined life or inspection requirement and design the warehouse method accordingly. Do not buy years of supply merely because the part is inexpensive; unusable inventory does not improve service.
Create a transparent starting model. For each part, list the launch quantity, installed-base assumption, criticality, planned service method, replenishment lead time, minimum order, compatibility, and uncertainty. Use scenarios rather than false precision: a base case, a plausible higher-demand case, and a response trigger. Reserve stock for warranty commitments separately from normal sales if needed. The approved plan should name who can release parts and how consumption is recorded.
Real field data should replace assumptions quickly. Review requests by model, batch, age in service, country, symptom, confirmed failure mode, and final disposition. A "replacement sent" count can hide setup problems, shipping damage, lost hardware, or parts exchanged without diagnosis. No-fault-found events are still useful when recorded accurately. Watch service levels and shortages as well as stock consumption; zero usage may mean no failures, poor awareness, or an inaccessible process.
Forecast demand in time buckets. Launch stock must cover the interval before the first replenishment can arrive, while later orders can respond to evidence. Include expected growth and a service buffer, but keep the uncertainty visible. A spreadsheet should show opening stock, receipts, issues, reserved quantity, available balance, reorder point, and compatibility status. Separate consumable packaging or installation accessories from warranty spares so frequent sales use does not hide service exposure.
Repair yield affects consumption. If an authorised centre replaces a controller during diagnosis and later returns the original to usable stock, gross issues and net use differ. Some parts may be scrapped after opening or require paired replacement. Record the service outcome and whether removed items are quarantined, analysed, repaired, or discarded. Do not circulate an unknown returned electrical component as a spare. The forecast should use parts that actually restore approved products, not every item that left the warehouse shelf.
End-of-life planning begins before discontinuation. Estimate how long products may remain in service under the company's support policy, identify last-buy decisions, and protect documentation and tooling needed for compatible parts. When a revision is superseded, decide whether a backward-compatible spare exists and validate any conversion kit. Communicate honestly when support changes. Excess old stock and shortages of a tiny unique mount are both easier to avoid when engineering, sales, and service share the same lifecycle forecast.
Commercial terms affect the plan but should not hide service risk. Decide who owns warranty stock, which party pays urgent freight, how distributors report use, and whether parts may be sold outside the authorised repair route. Include tax, customs, and local return constraints where relevant. A theoretical central stock pool is little help if no process can release a part quickly across borders. Test the ordering path with a sample service case before customers depend on it.
Forecast accuracy should be reviewed honestly. Compare predicted and actual use by part and period, but explain changes in installed base, lead time, campaigns, or product revision. A forecast that overstates one month may still have protected service during a delayed shipment. The aim is not a perfect percentage; it is an available, compatible part with controlled inventory and a clear decision when evidence changes. Document adjustments so repeated optimism or overstocking can be corrected.
A mature spare-parts forecast is a living service model. It protects customers from long waits without filling shelves blindly. Installed base, part criticality, lead time, approved repair policy, revision compatibility, storage limits, and early evidence all contribute different pieces. By reviewing them together, a distributor can hold the simple parts that restore normal use, escalate safety-critical work correctly, and adapt stock as the market teaches what the launch spreadsheet could not know.
