Most parts organisations measure fill rate somewhere that flatters them. A central warehouse reporting high availability is reporting on its own shelves — and the observation that matters is that even where headline fill rates look healthy, the customer-facing figure is very often considerably worse, because of the distribution, dealer and service networks sitting in between. That gap is where the business actually lives. A technician with a vehicle on a lift does not experience your central number; they experience whether the part is on the shelf behind them, and if it is not, whether it arrives today or on Thursday. Managing the network means measuring at the counter and working backwards, which is a different exercise from optimising each echelon on its own terms. See it on your own data — bring a month of dealer order lines and we'll show where the gap opens.
2026 GUIDE · PARTS DISTRIBUTION
OEM Parts Distribution Network Operations
Central warehouse
The parts counter
Same network, two very different numbers — and only one of them is the customer's experience
Where the Gap Opens
Four places between your shelf and the technician's hand. Each one is measurable, and most organisations measure none of them.
01
Central holds it, the region does not
A line filled from central rather than regional is filled — and it is also a day slower and a leg more expensive. Counting it as a success at the central level hides a stocking decision that was wrong at the regional one.
02
The region holds it, the dealer did not order it
Availability upstream is irrelevant if the dealer's own stocking profile never included the part. This is the largest single gap in most networks and it is a replenishment design problem rather than a supply one.
03
The dealer ordered it and missed the cutoff
Orders into manufacturer systems run against a daily cutoff window — commonly early to mid afternoon for next-day programmes. A line placed twenty minutes late is not a supply failure, but the technician cannot tell the difference.
04
Nobody ordered it at all
The part was unavailable, the dealer sourced it elsewhere, and no demand signal ever reached you. That line does not appear in your fill rate as a miss — it does not appear at all, which is the subject of its own section below.
Note the pattern. Only the first two are visible in a conventional availability report, and the fourth is invisible by construction. A network measured only at its own echelons will report improvement while the counter experience stays flat — which is exactly the disconnect the customer-facing figures keep revealing.
Benchmarks Worth Measuring Against
Three published figures. The third is the one that reframes the conversation, because it is a gap rather than a target.
90%
First-time off-the-shelf fill rate — the commonly cited baseline benchmark for a dealer parts operation
95%
Same-day pick-up fill rate — treated as the stronger availability target rather than the baseline
95.6% vs 67%
Availability on manufacturer-managed replenishment against parts the dealer chose to manage themselves — reported under one long-running programme
Read that third figure carefully
It is not a comparison between two suppliers or two warehouses. It is a comparison between two decision-making methods applied inside the same dealership — parts under managed replenishment against parts the same team chose to stock on judgement. Nearly thirty points of availability separate them, and the difference is not effort or intent. It is whether the reorder decision was driven by a system or by memory.
Availability and Inventory Move Together, Not Against Each Other
The objection that stops most replenishment programmes is that higher service must mean more stock. The reported experience runs the other way.
Availability rose
To the mid-nineties on managed lines, against roughly two-thirds on self-managed ones — a difference large enough to change what the service department can promise a customer.
Inventory fell
Participating dealerships reported reductions of 15 to 50 percent, after an initial spike on joining. The stock was not too small — it was in the wrong parts.
Emergency orders fell too
Reductions in emergency and special orders came alongside both. Three measures improving together is the signature of a planning problem being solved rather than a budget being increased.
There is a second, less comfortable finding in the same history worth stating. Some of the supply fluctuation being corrected was caused by the manufacturers themselves — discounting and awards programmes encouraged parts managers to load up at particular points in the year, distorting demand signals that the same organisations then tried to forecast against. Incentive design is part of network design.
Central
↓
Regional
↓
Dealer
↓
Counter
Which echelon is your fill rate measured at?
If the answer is any of the first three, the number is describing your network to you rather than describing your network to the customer. Bring a month of dealer order lines to a 30-minute session and we'll rebuild the measurement from the counter backwards — showing which echelon each miss actually originated at.
Emergency Orders Are a Diagnostic, Not a Cost Line
They get managed as an expense to be squeezed. They are more useful read as a measurement of how much of your demand you failed to anticipate.
Below 5% is achievable
Operations that link parts ordering to the service schedule — rather than ordering what they ran out of — report emergency order rates below five per cent, described as well under the industry average. The mechanism is a pull signal from booked work instead of a push reaction to an empty bin.
Campaigns are predictable demand
Recall work, technical bulletins and scheduled maintenance campaigns are known events. You know the campaign is coming, which parts it requires, and roughly how many vehicles in the parc are affected. Any campaign part that arrives as an emergency order was forecastable and was not forecast — which makes campaign-driven emergency lines the easiest category to eliminate entirely.
Zero is the wrong target
The aim is not to eliminate emergency ordering but to replace reactive ordering with anticipation. Some genuine urgency is legitimate, and squeezing the line without fixing the planning simply moves the failure onto the technician.
The Demand You Never See
The most damaging blind spot in a parts network, and the reason forecasts drift away from reality over time rather than toward it.
The substitution happens quietly
When a part is not available, the dealer does not stop working. They source elsewhere — an aftermarket supplier, an inter-dealer trade, another rooftop in the group. The vehicle gets fixed and the customer is satisfied.
And the demand disappears
That line never reaches your order history. It is not recorded as a miss, a backorder or a lost sale — it simply does not exist in your data. Channel-level demand gaps of exactly this kind are described as invisible until the installed base is connected to planning.
So the forecast trains on filtered data
Next period's plan is built from orders that were placed, which are by definition the ones you could fill. Parts you are bad at holding generate less recorded demand, which justifies holding fewer of them. The blind spot reinforces itself, quietly, every cycle.
Forecast forward, not backward
The alternative described is treating demand as a function of what equipment exists, where it sits in its lifecycle and what it will consume next — positioning what will sell rather than replenishing what sold. Connect the installed base to planning and forecasts become equipment-driven, obsolete stock becomes identifiable rather than merely suspected, and fill rate targets can be set by category instead of applied uniformly across every part number.
Vehicle Off Road Is a Different Product
Treating urgent lines as ordinary lines with a faster shipping option understates what is at stake.
The vehicle cannot be driven
That is the definition, and it is why every manufacturer operates a programme to move parts to the dealer as quickly as possible in that situation. It is a distinct service level, not an expedite flag on a normal order.
Days accumulate into legal exposure
Off-road days add up, and they can bear on lemon law thresholds. A parts availability failure on the wrong vehicle stops being a service inconvenience and becomes a legal and buyback question — which is a category of cost that never appears in a distribution budget.
So measure it separately
Off-road lines need their own fill rate, their own cycle time and their own escalation path. Blending them into the general figure hides the only lines where hours genuinely matter.
Designing the Echelons
Five decisions that determine whether the network can deliver what the measurement demands.
Stock by movement, not by echelon habitFast movers close to the counter, slow movers consolidated upstream. The test is not what each site has historically held but what the next-day promise requires it to hold.
Set targets by category, not uniformlyA uniform fill target across every part number over-invests in slow lines and under-invests in the ones that stop a vehicle. Competitive category and consequence should both influence the target.
Design around the order cutoffDaily ordering windows into manufacturer systems govern whether a line makes the next-day run. Cutoff discipline is a service lever that costs nothing, and missed windows are indistinguishable from stockouts at the counter.
Give obsolescence a defined routeReturn programme, transfer to another rooftop, aftermarket liquidation, or write down. Recovering forty to fifty cents on the dollar beats holding a part to zero value and writing it off later — and an unmanaged no-mover list distorts every stocking decision around it.
Pull demand from booked workLinking parts to the service schedule converts ordering from a reaction into an anticipation, and it is the change most strongly associated with low emergency order rates. Campaign volumes belong in the same signal.
The Measurement Set
Seven figures. Reported together they describe a network; reported individually they mostly describe a warehouse.
First-time off-the-shelf fill, at the counterThe headline number, measured where the customer stands rather than where the stock sits.
Same-day fillThe stronger availability measure, and the one that reflects whether the vehicle leaves today.
Fill source by echelonWhich lines were filled locally, regionally and centrally. This is what turns a single percentage into a stocking decision.
Emergency order rateAs a share of lines, tracked against the sub-five-per-cent figure reported by operations pulling demand from booked work.
Off-road lines and daysSeparately from everything else, with their own cycle time — because their cost profile is different in kind.
Missed cutoff linesOrders that failed the daily window. A free service improvement hiding inside a stockout statistic.
No-mover value and dispositionWhat is not selling, how much it is worth, and which of the four exit routes each line is on.
Twenty minutes on a month of order lines and you will know where your fill rate actually breaks
On a working session we'll take dealer order lines into Fleet Rabbit, rebuild fill rate measured at the counter rather than at the warehouse, attribute every miss to the echelon it originated at, split emergency and off-road lines out of the general figure, and identify the campaign-driven demand that should never have arrived as urgent. You keep the analysis either way — and for most networks it is the first time the counter number and the central number have been in the same view.
Frequently Asked Questions
Our fill rate looks good. Why do dealers complain?
Because the two numbers are measuring different things. Even where headline fill rates appear healthy, customer-facing rates are very often considerably worse, and the difference is created by the distribution, dealer and service networks in between. A line filled from central instead of regional counts as a success upstream and arrives a day late downstream. Measure at the counter and attribute each miss to the echelon it originated at, and the complaint usually stops being mysterious.
What should we be targeting?
Commonly cited dealership benchmarks put first-time off-the-shelf fill at 90% as a baseline and same-day pick-up fill at 95% as a stronger target. Treat those as orientation rather than as a specification, since the right figure varies with department mix, stock profile and market. More useful than either is the reported gap between managed and unmanaged replenishment inside the same dealership — around 95.6% availability on managed lines against roughly 67% on self-managed ones.
Won't better availability mean carrying more stock?
The reported experience is the opposite. Dealerships joining managed replenishment programmes saw an initial inventory spike, then reductions of 15 to 50 percent, alongside higher availability and fewer emergency and special orders. Three measures improving together indicates a planning problem being solved rather than a budget being increased — the stock was not too small, it was in the wrong parts.
How low can emergency orders realistically go?
Operations that link parts ordering to the service schedule report rates below five per cent, described as well below the industry average, by generating a pull signal from booked work rather than reacting to empty bins. Zero is the wrong target though — some urgency is legitimate. The category worth eliminating entirely is campaign-driven: recall work, technical bulletins and scheduled maintenance are predictable events where the parts and affected volumes are known in advance.
What is the demand we cannot see?
Substitution at the dealer. When a part is unavailable the dealer sources it elsewhere — aftermarket, inter-dealer trade, another rooftop — and that line never reaches your order history. It is not recorded as a miss or a lost sale; it simply does not exist in your data. Next period's forecast then trains on the orders you could fill, so parts you are poor at holding generate less recorded demand, which justifies holding fewer. The blind spot reinforces itself every cycle.
Why measure off-road lines separately?
Because the consequence is different in kind. An off-road vehicle cannot be driven, which is why manufacturers operate dedicated programmes to move those parts fastest. Off-road days accumulate and can bear on lemon law thresholds, so a parts failure on the wrong vehicle becomes a legal and buyback exposure rather than a service inconvenience. Blending those lines into the general fill rate hides the only ones where hours genuinely matter.
Where should we start?
By recalculating fill rate at the counter and attributing every miss to its origin — regional stocking, dealer profile, missed cutoff, or a line that was never ordered at all. That single change reframes the whole network conversation, because it separates supply problems from replenishment design problems, and the two need completely different responses. Book a session with a month of order lines and we'll build it with you.
Measure at the Counter, Fix at the Echelon
90% baseline95% same-day<5% emergency
Attribute every miss to where it originated, keep off-road lines out of the general figure, pull demand from booked work and campaign volumes, and remember that the lines you cannot fill are quietly disappearing from the data you plan with.
Benchmarks are drawn from publicly reported industry figures and vary by market, franchise and operating model — treat them as orientation rather than as targets for your own network
September 7, 2026By Alex Rowan
All Blogs
