Walk into any UK Authorised Treatment Facility and you will see the same problem on the shelves: stock listed against a single model, a single year, a single side. The wing mirror you took off a 2016 Astra J is filed as "Astra J 2016 LH". The fact that the same mirror was fitted to four other variants across two model years is invisible to the search.
What the part actually is
A wing mirror is not a wing mirror. It is a housing, a glass, a base, a motor, a heating element, an indicator repeater, a wiring harness with a particular pin count, and a colour code. The same OEM part number ships against any vehicle whose specification matches all of those — which, for a volume model, is usually several variants spanning several years.
The Astra J driver-side mirror with power-fold and heating, in primer, is the same Vauxhall part number across pre-facelift and facelift cars, across hatch, estate and saloon, across multiple engine variants. One number, five model rows on the fitment side.
A donor vehicle is a fan-out, not a list
When a dismantler depollutes and strips a donor vehicle, the items coming off the shell are not really "Astra J 2016 hatch" parts. They are parts that fit everything sharing their specification. The graph turns one donor into matches across the whole fitment fan.
In practice: a 2016 Astra J hatchback comes in. The dismantler photographs and lists ~80 sellable items. Against a flat catalogue, those 80 items are findable only by buyers searching exactly "Astra J 2016 hatch". Against a graph, each item resolves to every vehicle it actually fits — five model variants across two years for the average exterior part, four to six fuel/gearbox combinations for the average mechanical part.
What it does to inventory turn
Inventory turn at an ATF is brutal. Anything not sold inside ninety days is, statistically, never going to sell at anything like the listed price. Widening the search surface for each part — by being correctly cross-referenced from day one — is the single biggest lever on that ninety-day window.
- Flat catalogue search surface
- 1 variant × 1 year × 1 side
- Graph search surface (typical exterior part)
- 5 variants × 2 years × 1 side
- Multiplier on impressions
- ~10×
- Observed effect on 90-day sell-through
- 1.6×–2.4× depending on category
Those numbers are why the dismantling side of the business changes shape once it joins a real cross-reference. The same shelves, the same stock, the same staff. More of it sells, faster.
It cuts the other way too
The reverse is just as important. When a repairer searches a part number for a 2014 Insignia, a graph lookup can surface the matching used part from a 2015 donor sat in a Birmingham yard, because the OEM number is shared. A used part in stock, twenty miles away, today, beats a new part on a four-day order from a national distributor. The repairer wins, the fleet wins, and the ATF turns inventory it would otherwise have weighed in.