Guides

Every benchmark target should tell you where it came from

Somewhere in every benchmarking system there is a number in a column headed "target", and a dealer being measured against it. Ask where that number came from and the answer is often a shrug, or a sentence about experience, or silence.

That is a problem, because the six ways a target can be set are not equally trustworthy, and a dealer being held to one deserves to know which kind he is looking at.

This guide sets out the six, explains why the difference matters operationally, and describes how Apex20 handles it.

The six ways a target gets set

1. Peer-derived. Calculated from the actual performance of the stores in the group. The top quartile of the room on that line, or the group average, or some other function of real numbers that real stores produced in a real month. This is the strongest kind of target, because it is demonstrably achievable: somebody in the room did it.

2. Researched. Derived from published or purchased data outside the group. Useful when the group is small or when a line has too few reporting stores to be meaningful. Weaker than peer-derived, because the population behind it is not your population, and it can go stale silently.

3. Expert judgement. Someone experienced looked at the line and set a number. This is more defensible than it sounds when the person is genuinely expert and the reasoning is written down. It is indefensible when neither is true, and it is the category most often disguised as one of the other five.

4. Imported. Carried over from a prior system when a group moved. Perfectly reasonable as a starting point and dangerous as a permanent state, because the reasoning behind the original number did not come across with the number.

5. Manual. Typed in deliberately by the group, usually because the group agreed on a goal. Legitimate and often the most motivating kind, provided everyone remembers it is a goal the group chose rather than a fact about the world.

6. Override. A person changed a target that the system had already set. The important thing about an override is not that it happened, it is that the record should say it happened, who did it and when.

Why the difference is operational, not academic

Consider two stores looking at the same target on the same line.

The first is a metro store with a large service drive, a full technician complement, and a customer base that turns over quickly. The second is a rural store with four bays, one of everything, and customers who have been coming for twenty years.

If the target on that line is peer-derived from a group that contains both kinds of store, it is a useful stretch for both and a fair yardstick for neither in isolation. If it is researched from a population of metro stores, it is quietly unfair to the rural store, and the rural dealer will know it is unfair without being able to say why, which is worse than knowing.

The rural dealer's proper response to a researched target is not to dismiss it. It is to ask what population it came from, and then to decide how much weight it deserves. That conversation is only possible if the system says where the number came from.

Here is the practical consequence. A dealer who trusts a target will chase it. A dealer who suspects a target will explain it away, and every subsequent conversation about that line becomes an argument about the benchmark rather than about the operation. Provenance is what keeps the conversation on the operation.

Why the number without its origin is worse than no number

An unexplained target does three things, all bad.

It transfers authority to whoever set it, permanently and invisibly. Nobody can challenge a number whose basis is unknown, so nobody does, and it becomes fact by survival.

It hides staleness. A researched target set from a data set that is four years old looks exactly like one set last month. Only the provenance record reveals the difference.

It makes disagreement look like excuse-making. When a dealer says "that target does not fit my store", a system with provenance turns that into a checkable question. A system without it turns it into a character assessment.

What Apex20 does about it

Every target in the Apex20 book carries a badge saying which of the six it is: peer-derived, researched, expert judgement, imported, manually set, or overridden. The badge is on the line, next to the number, not buried in documentation. You can always ask why a target is what it is, and get an answer.

The member-facing basis of the book is deliberately narrow and deliberately stated:

That is the whole basis. The book carries no external industry benchmark. We do not publish one, we do not imply one, and we do not quote one, because we cannot show you the population behind it and a benchmark you cannot audit is a benchmark you should not be measured on.

The same principle runs through the rest of the book. Click any line and the portal shows your own arithmetic, your own gross divided by your own sales, sourced back to the line on the statement you uploaded. The target is treated the same way as the result: it shows its work.

What to ask your current provider

If you are already in a group with a book you did not build, these questions are fair and worth asking:

A provider who can answer all five is doing the job properly. A provider who cannot may still be producing a useful book, but you now know which numbers to weigh lightly.

Why this is hard to copy

Provenance is not a feature you add in a sprint. It has to exist at the point a target is created, survive every import, every recalculation and every period lock, and be carried all the way to the line the member reads. Systems that did not do it from the beginning generally cannot retrofit it, because the origin of the existing numbers is genuinely lost.

We built it in from the start because the question "where did that come from?" is the question a dealer asks about every number he does not like, and the answer decides whether the meeting is productive.

If you want to see a target and its badge on a real line, call and we will walk you through it.