Some gaps need correction, not diagnosis

Some requirements are non-negotiable. A failed safety control, legal obligation, authorised-repair condition or customer-data rule requires prompt correction and verification. The cause can still be investigated, but the breach must not be treated as an acceptable performance variation. The diagnostic method below concerns differences in commercial and operational performance, including where a dealer follows the documented standard yet underperforms.

Standards describe the expected process. But they dont heal and Variances are still important.

A standard can specify how quickly a lead should be handled, which steps must occur during service reception or how an ageing vehicle is escalated. An audit can show whether the steps were followed. Neither establishes that non-compliance caused a commercial outcome.

A dealer may follow the enquiry process correctly but receive a very different mix of leads from another site. A workshop may meet reception standards but struggle to retain customers because parts are unavailable or technician capacity is constrained. A dealer may have unusually old stock because allocations and local demand diverged months before the field visit.

First, compare like with like

A raw network ranking can confuse opportunity with execution. Territory potential, customer mix, model availability, sales channel, workshop throughput and local competition all affect outcomes. Comparison is useful only when those differences are visible and the underlying metric has one definition across the network.

This does not require an elaborate statistical model for every field decision. It requires discipline about the denominator, the comparison group and the time period. A lead-to-sale rate is hard to interpret if one site counts all digital enquiries and another counts only qualified customer conversations. Service retention can mislead if the eligible vehicle population, vehicle age or fleet mix differs materially.

Start with the spread of results among genuinely comparable outlets rather than the national average. Where a fair comparison is impossible, say so rather than presenting a ranking as evidence of management failure.

Then separate the causes into five layers

A practical first diagnosis should test five possible sources of variance:

The five layers are a working diagnostic, not a claim that every problem has one clean root cause. Layers interact. The purpose is to avoid prescribing a dealership action for a constraint the dealership cannot control.

A variance hypothesis is more useful than an action list

Consider an illustrative example. Two otherwise comparable service outlets have materially different rates of completing promised work on time. The standard response is more service-adviser training. A stronger first step is to test the mechanism: do delays begin in appointment booking, work authorisation, technician allocation, parts replenishment or customer approval?

The field visit should end with a hypothesis that could be proved wrong, the evidence needed to test it, one accountable intervention owner and the expected sign of improvement. Ten activities with no causal explanation do not yet make a performance plan.

Do not copy the top dealer until the mechanism is visible

Rankings are good at finding questions and poor at giving transfer instructions. The top site may benefit from territory, product mix, a legacy customer base, spare capacity or one individual whose practice is not embedded in the business.

Before rolling a practice out, establish which behaviour changed which result, under which conditions, and whether it holds when the person, period or site changes. Transfer then means changing a routine, role, tool or decision rule. Circulating a best-practice presentation rarely changes any of them.

Make the intervention part of the management rhythm

The next step is a routine that connects diagnosis to action. A useful dealer-performance review can be organised around five questions:

This changes the job of network field teams. Their value moves from checking compliance and distributing a centrally prepared improvement list to distinguishing causes, securing cross-functional decisions and testing whether interventions travel from one outlet to another. That requires comparable data before the visit, and the authority to leave a site without prescribing a local action when the evidence points upstream.

The importer cannot diagnose away its own responsibility

The most expensive misdiagnosis is classifying every gap as dealer execution because that is where it is measured. The point of observation and the point of causation are often different. Allocation, targets, lead routing, warranty decisions and system interfaces are set above the dealer, and their effects surface in the dealer's numbers.

at.Pointe's published importer-transformation case (at.Pointe engagement, anonymised) describes this pattern. Importer, market and retail processes had grown separately, management data could not always be trusted enough to drive action, and systems did not connect cleanly. The redesign introduced common KPI definitions so that dealer and importer teams worked from a shared performance logic, with escalation routines designed around it. The case shows that upstream operating design shapes dealer results. It does not show that any particular diagnostic method improves profit.

An effective network-performance system therefore needs an escalation route for causes that require importer or OEM changes. Without one, field teams can identify a problem accurately and still have no authority to solve it.

The management decision

Standards remain necessary. They set minimum operating expectations, protect the customer promise and make processes repeatable. The next improvement may not be a thicker standards manual or another network dashboard.

It is more likely to be a better way to compare outlets, investigate why they differ and direct management effort to the organisation that can change the outcome. That is how a network moves from seeing variance to managing it.