Service Parts Planning: The Eight Views a Parts Business Runs On

A service parts operation does not have an inventory problem. It has eight problems wearing one inventory costume. A distributor can hold AED 40m of spare parts across four warehouses and still fail to fill a same-day order for the bearing a field engineer needs on site by noon. It is never too little inventory — it is too much of the wrong inventory, and not enough of the three lines a customer is waiting on right now.

The reason a single stock report never fixes anything is that service parts performance is really eight separate questions, each owned by a different person, each measured differently, and each invisible to the others. The parts counter sees stock on hand. Planning sees a reorder report that ran overnight. Finance sees a working-capital number at month-end. Procurement sees supplier lead times in a spreadsheet. Nobody sees the same picture at the same time — so the reconciliation happens in Excel, after the window to act has already closed.

Put the eight views on one canvas and the trade-offs stop being arguments and start being decisions. What follows is a walk through the eight views an aftermarket parts operation genuinely runs on — built as a single Microsoft Fabric and Power BI model over an iScala item master, with the data integration landing every source into one governed layer.

Three of the eight views answer the only question a customer actually cares about: did the part arrive, on time, in full? The On-Time Delivery view tracks OTD and OTIF against a 95% target, rolling twelve months — then does the thing a spreadsheet never does. It root-causes the misses. Late lines break down by reason (supplier delay, backorder, transport and customs, documentation, picking error) and by customer segment, so an OEM-contract account slipping below SLA is visible before the account manager hears about it on a call.

The Customer Service view sits next to it: line fill rate against target, open backorders, service-order queue, and a backorder worklist ranked by impact, so the counter team works the lines that matter rather than the loudest caller. And Service Level Tracking takes the same signal to the contractual layer — SLA attainment, breach count, mean time to resolve, and attainment by warehouse. When Riyadh is breaching and Dubai is not, you see it, and you see whether it is a stocking-policy problem or a supplier problem, not just a red number.

Two views turn the service story into a money story. Inventory Health is where the working-capital conversation gets honest. Total inventory value, turns against a 4.0x target, and — the number most parts businesses avoid — the share of value locked in excess and slow-moving stock. The days-of-supply distribution sorts every SKU into cover bands, so you see the two tails at once: the parts about to stock out, and the parts you will still be holding in three years.

Inventory Optimisation turns that diagnosis into a number finance can bank. It compares current reorder points against a model-recommended reorder point for every high-value part, quantifies the safety-stock capital sitting in each ABC class, and puts a figure on the working capital you could release by redeploying excess. This is the view that pays for the engagement — not by buying less across the board, but by holding the right mix. A parts operation running a blanket "keep three months of everything" policy is financing its C-class shelf-warmers with the capital its A-class fast movers need.

Two more views drive what gets bought, and when. Demand Planning is the honest one. Forecast accuracy (1 minus MAPE) against an 80% target, forecast bias, and — critically — how many SKUs are actually forecastable at all. Spare-parts demand is intermittent and lumpy; a large share of any parts catalogue has a demand coefficient of variation high enough that no statistical forecast will ever be reliable. The view names those parts explicitly, because pretending to forecast them is how planners lose trust in the whole model.

Supply Planning closes the loop back to procurement: open purchase orders and committed spend, projected incoming supply against forecast demand across the next twelve months, supplier lead times, and a supplier scorecard ranking on-time performance and open commitment. When a stock-out-risk part already has a PO in flight, the recommended action is expedite, not raise another order — a distinction that quietly saves a fortune in duplicate emergency buys.

The eighth view is the one that makes the other seven worth reading: Item Master Governance over the iScala data. Completeness against a 98% target, records with missing lead times, missing suppliers, missing units of measure, and duplicate-risk parts — with an exception worklist a data steward can actually work through.

This is not a housekeeping afterthought. A missing lead time does not produce a wrong answer; it produces no answer — the reorder-point maths silently fails for that part, and it drops out of every replenishment run without anyone noticing. Duplicate SKUs split one part's demand history across two records and wreck the forecast for both. Every clean number on the other seven pages depends on this one being addressed first. It is the least glamorous view and the most important — which is exactly why it usually gets skipped, and exactly why the platform breaks when it does.

The eight pages are one connected model, not eight dashboards. The interactive report steps through On-Time Delivery, Inventory Health, Customer Service, Demand Planning, Inventory Optimisation, Supply Planning, Item Master Governance and Service Level Tracking — the same Power BI-style report, rendered with live charts. It is a demonstration model built on synthetic data, representative of a mid-market parts hub with around 120 tracked SKUs across Dubai, Jebel Ali, Abu Dhabi and Riyadh. The structure, KPIs and logic are exactly what we build in production; the numbers are illustrative.

In a live engagement these views refresh against your ERP on a governed schedule — not overnight, and not by hand. That is the difference between a picture of last week and a picture of this morning.

No report fixes a parts business on its own. The forecast has a ceiling, and it is lower than vendors admit. Intermittent spare-parts demand means a meaningful slice of your catalogue is buffer-driven, not forecast-driven. For those parts the right answer is a well-set safety stock and a fast replenishment loop — not a cleverer forecasting model. A view that pretends otherwise is selling confidence it does not have.

Garbage in the item master, garbage on every page. If lead times are missing and parts are duplicated in iScala, the reorder points are wrong and no dashboard will save you. The governance view sizes that problem; fixing it is data-stewardship work, and it comes first — not a two-week job you can skip on the way to the pretty charts.

And a view is not an action until someone owns it. The report surfaces the stock-out-risk list and the capital-release opportunity. It does not raise the PO or approve the write-down. The value lands when the daily worklist has an owner and the automation layer raises routine replenishment on its own — which is a separate build on top of this one. Reserve automation for the bounded replenishment decisions and keep humans on the judgement calls; over-reaching here is how trust in the whole model collapses.

If you run a service parts or aftermarket operation and your answer to "what is our fill rate right now, and what is it costing us in held capital" involves exporting three reports and reconciling them by hand — that gap is the problem, and it is fixable. The point of putting these eight views on one governed foundation is not the visuals. It is that the person accountable for service and the person accountable for working capital finally read from the same number, at the same time, early enough to act.

And it builds in sequence, not as a big bang: a six-week Discover and Foundation build stands up the governed data layer and the first live views on real data; optimisation, supplier scorecards and automation are added on top as the item master matures. First value in six weeks, compounding from there. In the service parts operations we deliver, that shift typically releases 15 to 30% of the working capital tied up in the excess and slow-moving tail — without touching the fast movers.