A worked example: warranty for an auto parts distributor

This is a worked example, not a customer account. No distributor described here exists, no outcome below was measured anywhere, and every figure is an assumption you should replace with your own. What transfers is the structure of the argument, not the totals.

The archetype

A regional distributor of automotive parts supplying independent garages. Several hundred trade accounts, next-day delivery, a catalogue running from consumables to serialised units — alternators, starters, turbochargers, injectors, clutches, remanufactured assemblies.

The distinguishing feature of this business is that the person who claims the warranty is almost never the person who owns the vehicle. A garage buys the part, fits it, and charges the motorist for parts and labour. When the part fails, the garage claims — and the garage also wants its labour back.

That gap between purchaser, fitter and owner sits underneath every problem below.

The assumptions

Assumption	Value
Trade accounts	300 garages
Serialised, warrantable parts sold per month	4,000
Claim rate	2.5% of warrantable parts
Share of claims that are fitment-related, not defects	30%
Average part cost to the distributor	120
Remanufactured units sold per month	1,500
Core charge per remanufactured unit	45
Cores never returned	12%
Share of parts fitted to vehicles in commercial use	20%
Time spent on a contested claim today	40 minutes

Currency is whatever you work in. From the above: 4,000 × 2.5% = 100 claims a month, of which roughly 30 are contested on fitment grounds.

Who is the customer, exactly?

Worth settling before anything else, because the whole workflow follows from the answer.

Party	Holds	Wants	Knows
Distributor	The commercial relationship, the terms	To pay only for genuine defects	What was sold, to whom, and when
Garage	The account, the fitting labour	Part replaced and labour covered	What was fitted, to which vehicle, and how
Motorist	The vehicle and the failure	A working car	The symptom and the mileage
Manufacturer	The underlying warranty	Evidence before crediting	Batch and production data

Every fact needed to settle a claim exists. It is distributed across four parties and never assembled in one place. That is the actual problem, and it is not solved by tightening the terms — it is solved by capturing the facts at the moment each party has them.

Problem 1: the claim arrives from the fitter

The garage phones or emails: part failed, we have fitted a replacement, please credit us. What they typically supply is an invoice number and a part number.

What is missing is everything that identifies the specific unit and its working life: the serial, the vehicle it went on, the date it was fitted, the mileage then and now.

Without that, the distributor is settling a claim about a part number rather than a part. It cannot tell whether this is the unit it sold, whether the unit has already been replaced once, or whether the failure happened after three weeks or after eighteen months.

Problem 2: fitment errors arrive as defect claims

This is the expensive one, and it is rarely dishonest. A turbocharger that fails because the oil feed was not cleaned, a clutch that burns out on an unresurfaced flywheel, an injector contaminated by a fuel system nobody flushed — each of these is a part that genuinely failed and genuinely was not defective.

The distributor knows this class of failure exists. What it cannot do is demonstrate which claims belong to it, because it has no record of how the part was installed. In a dispute with a trade account it values, the commercial reality is that it concedes.

The record that changes this is a fitment record, captured by the garage at installation:

• Serial number of the unit fitted
• Vehicle identifier and mileage at fitting
• Date and the fitting technician
• Confirmation of the prerequisite steps for that part — oil feed cleaned, flywheel resurfaced, system flushed, sensor replaced

That last item is the one that does the work. It converts an argument about what probably happened into a question about a field somebody either completed or did not, at a moment when completing it honestly costs nothing.

It also gives the garage something: a fitment record is evidence for them when the installation was correct, which is most of the time.

Problem 3: core returns leak quietly

A remanufactured unit is sold with a core charge, refunded when the old unit comes back. Cores go missing — left on a workshop floor, scrapped, forgotten on an account that has since paid its balance.

Using the assumptions: 1,500 remanufactured units at a 12% non-return rate is 180 uncollected cores a month, or 180 × 45 = 8,100 a month in charges that must be chased, written off, or quietly absorbed.

The problem is not the policy. It is that nobody holds a live list of which specific units have an outstanding core, by account and by age. When the core obligation attaches to the serial at the point of despatch, that list generates itself, and chasing becomes routine rather than an annual clean-up.

Assume disciplined chasing recovers half of what is currently lost: 90 × 45 = 4,050 a month. Your own non-return rate and recovery rate will differ; the point is that the figure is currently unknown to most distributors, which is the first thing to fix.

Problem 4: commercial use on consumer terms

Suppose the distributor offers 24 months on parts fitted to privately-owned vehicles and 12 months or a mileage cap on vehicles in commercial use — taxis, delivery vans, fleets. Entirely reasonable, and near-impossible to enforce if nobody records which is which at the point of sale.

The default, in practice, is that everything is treated as consumer use, because that is what the garage will assert and there is nothing to contradict it.

The arithmetic, with stated assumptions: 20% of 4,000 parts is 800 units a month going onto commercial vehicles. Assume duty cycle doubles the claim rate to 5%, giving 40 claims a month from that population. Assume half of them would fall outside the commercial term on age or mileage: 20 × 120 = 2,400 a month paid against terms that did not apply.

Capturing use type is one field at the point of sale. Enforcing it requires that the terms in force at that sale are the ones tested at claim time — which means versioned terms bound to the unit, not a current PDF. That distinction, and how to write terms that survive it, is covered in how to write warranty terms.

Putting the redesigned flow together

Moment	Record created	What it later settles
Despatch to garage	Serial, account, date, core obligation, terms version	Whether the unit is ours; which terms apply; what core is owed
Fitting	Vehicle, mileage, date, technician, prerequisite checks	Whether fitment was correct; when the clock started
Claim	Serial, symptom, mileage now, prior claims on the unit	Coverage, duplicates, repeat failures
Core return	Serial matched against the obligation	Credit released automatically
Vehicle sold	Coverage follows the vehicle, not the account	Who may claim next

The last row is worth pausing on. If the remaining coverage on a fitted part belongs to the vehicle, then a change of owner is a warranty transfer — a logged event with a clear end and a clear beginning — rather than an informal assumption that whoever presents the invoice is entitled to claim.

The arithmetic, assembled

Monthly, on the assumptions stated above:

Line	Working	Monthly
Fitment claims conceded today	30 contested, 2 in 3 conceded, at 120	2,400
Fitment claims conceded with evidence	30 contested, 1 in 3 conceded, at 120	1,200
Difference		1,200
Core charges recovered	Half of 180 uncollected, at 45	4,050
Commercial-use claims correctly assessed	20 claims at 120	2,400
Illustrative total		7,650

And the time, separately: 30 contested claims at 40 minutes is 20 hours a month. If having the fitment record on screen takes that to 12 minutes, it is 6 hours — 14 hours returned, which is most of a working week across the year.

Two warnings about that table. First, it stacks four assumptions, so it is only as reliable as the weakest of them; the conceded-claims ratio is the one to challenge hardest, since it is the one nobody measures. Second, none of these are savings you can book — they are exposures you can now see and act on. The difference matters.

What this does not fix

• Garages still have to complete the fitment record. If it takes more than a minute on a phone, it will be skipped, and a skipped record is worse than none because it creates the appearance of evidence. Make completion a condition of the labour claim, not a favour.
• Genuine ambiguity remains genuine. A turbo can fail from a marginal oil feed and a marginal casting at the same time. Better records narrow the disputed zone; they do not abolish it.
• Non-serialised lines stay out of scope. Filters, pads, belts and fluids are batch-tracked at best, for the reasons set out in serial number tracking.

Where to start

1. Attach the serial and the terms version at despatch. No behaviour change outside your own warehouse.
2. Build the open-core list. Same data, immediate cash effect, no dependency on garages.
3. Add the use-type field at order entry. One question, asked of the account placing the order.
4. Introduce the fitment record last, because it is the only step that requires other people to change what they do — and make it worth their while by paying labour claims faster when it is complete.

Steps one to three are yours alone and pay for themselves. Step four is where the fitment disputes get resolved, and it will take a season of trade-counter conversation rather than a software rollout.

---

Warranlytics holds warranties against the serial number with the terms that applied at that sale, records service and fitment events on the same unit, and logs transfers of remaining coverage — so a contested claim opens with the history rather than requiring it to be assembled. See how a claim runs end to end, or look at the plans.