Perfect store execution: what happens when the standard itself is wrong

A perfect store programme sets the in-store standard for range, placement, price and promotion. Each store is then scored against it every cycle. The programme can fail in two ways. Either the standard was not met, or the standard was wrong. Compliance scoring finds the first one. Perfect store execution finds the second only when the standard is tested before it is issued. Many programmes never test it.

You will know the shape of it. The scorecard comes back in the nineties. The category is still flat. The field team is being asked to try harder at something that may not be the problem.

What a perfect store execution programme is

A perfect store programme turns the word ‘good’ into something a person can score in an aisle. Four things are usually named. Range is which products sit on the shelf. Placement is where they sit. Price is what the shopper is charged. Promotion is what is displayed around them.

Each store is scored against a picture of success. That is the drawn and written version of what the aisle should look like. The scoring runs store by store, cycle by cycle, so the programme produces a number for every store on every visit.

The split that matters is who owns which part. The standard is drawn centrally, by a category or customer marketing team. It is executed locally, by a field team standing in a store nobody consulted. The rest of this article lives in the gap between those two facts.

The approach has a citable origin. Bain & Company published it in October 2011 under the name Perfect Sales Execution. Bain described it as a scientific method for targeting the right stores and the right products, backed by a disciplined approach to supporting sales execution. The same 2011 article reported a survey of 120 consumer products executives. Ninety per cent put sales execution in their top five business priorities. Fewer than half felt their sales teams were working at full potential.

What a perfect store score tells you, and what it does not

A perfect store score measures one thing. It says how closely the store in front of you matches the picture of success it was issued.

Take a store that comes back at 95 per cent. That number tells you the shelf was built the way it was drawn. It says nothing about whether the drawing was right for that store, that fixture or that shopper.

The consequence turns up in the next review meeting. A compliance number gets read as a performance number. It is often the only figure in the pack with individual stores attached to it. A category that underperforms against a well-executed standard then produces an argument about effort. The plan is the one thing in the room nobody is testing.

That pattern is worth naming. A high score next to a flat category is not a contradiction. It is the programme answering the only question it measures. The question it does not measure stays open.

The same split exists one shelf at a time. For a question about a single planogram rather than a whole store standard, the piece on why planograms fail after rollout is the closer read.

Where a wrong standard comes from

A wrong standard is rarely a careless one. It is usually a standard nobody saw the way a shopper sees it.

Most standards are built in the same order. A range decision is made, space is allocated, and the result is signed off on a planogram. A planogram is a flat diagram showing which product goes where on a shelf. It then goes out to the network as the picture of success. Often nobody has stood in front of it in three dimensions.

Fixtures are the second source. A fixture is the shelving itself, and a bay is one section of it. A network usually holds several store formats. Bay widths, shelf heights and depths all differ between them. Draw a picture of success for the flagship format, issue it unchanged to the small format, and it cannot be built across much of the network. Those stores then score badly for a reason that has nothing to do with effort.

The third source is documented rather than asserted. Bain’s 2011 account describes companies using one globally standardised metric for measuring sales performance. Country managers were left to define assortment and positioning locally. A single metric laid over locally varying range and fixtures is how a standard ends up wrong somewhere.

None of this is visible in the score itself. It shows up as non-compliance, and what it costs when planogram compliance breaks down is a subject of its own.

How you would test a standard before it ships

Testing a standard asks two questions of it. Can it be built on the fixtures it is going to? And does it do the job it was drawn to do?

The sequence has three steps. Build the standard in a virtual copy of the store it will live in. Walk it the way a shopper would walk it. Then put it in front of shoppers and measure whether the product it was built around is found.

The first two steps answer the buildability question. A bay that is too narrow. A facing count that will not fit. A hero product sitting at ankle height in one format. A facing is one product front on the shelf, and the count is how many of them the standard gives that product. All of it is visible the moment the standard stands up. This work runs on the planogram files you already have, which can be brought in, changed and saved as versions.

The third step answers the effectiveness question. Shoppers can be put in front of the standard in a headset with eye tracking. They can also be sent through an online simulated shop with a survey afterwards. What comes back is whether the product the standard was built around was seen and chosen.

Keep the unit of test at the level of the whole standard: the picture of success, the fixture set, the format. Testing one planogram in one category is a narrower job with a narrower question.

Testing is a step, and a step costs time. What it buys is the subject of the next section.

What changes in the programme when the standard has been tested

The first change is to the number. A store scoring badly against a tested standard is a store problem. Nobody has to argue about whether the standard was achievable, because that was settled before it shipped.

The second change is to field time. A visit that opens with a negotiation about the plan is a visit spent on the plan. When the plan has been tested, the visit gets spent on the shelf instead. Mobile retail execution tools go on recording the same things; what changes is that the standard behind them holds up.

The third change is to the evidence. Three things now line up: a tested standard, a store walk that shows what the standard looks like, and geo-tagged photographs with compliance ratings coming back from the field.

That loop is what a retail execution strategy is for. It decides what gets settled before a cycle starts, and what gets measured while the cycle runs.

Be honest about the calendar. Testing sits before the cycle, not inside it. A programme that cannot fit it in before a reset should not try to test everything. Test the formats you are least sure about, which is usually the smallest and the newest.

What testing a standard does not fix

Testing settles the plan. It leaves a good deal untouched, and a programme that expects more from it will be let down.

It does not tell you whether the store had stock. An empty facing scores much like a wrong one. No simulation knows which of the two you are looking at.

It does not replace measurement in the field. A tested standard still has to be built in hundreds of stores, by people who were not part of the test.

It does not settle a range decision taken on commercial grounds. A standard can be drawn correctly around a range the category never wanted. Testing will only confirm that it can be built.

It does not fix a standard that is right and is simply not being executed. That is the first failure mode, and it needs field work rather than a simulator.

The case for testing is not that it lifts every score. It is that a wrong standard is paid for once per store, once per cycle. It keeps being paid until somebody looks at the plan again. The next standard your team issues is the cheapest one to test, because it has not been built anywhere yet.

Bring one contested standard from the current cycle to Storelab, with the fixture set it has to survive. Have it built and walked before the reset is locked. Send three things with the first message:

• The picture of success as it stands for that standard.
• The planogram files behind it.
• The list of store formats it has to work in.

Frequently asked questions

What is perfect store execution?

It is the work of setting an in-store standard, issuing it across a network, and measuring how well each store builds it. The term covers both halves: what good looks like, and whether it happened.

What does a perfect store score measure?

It measures how faithfully one store has built what it was issued. A high score means the build is close to the drawing. It carries no opinion about the drawing.

Can a store be fully compliant and still underperform?

Yes. Compliance is about the build, while sales come from what the build does for the shopper. A plan that was wrong for that store will not be rescued by doing it perfectly.

How do you test an in-store standard before rolling it out?

You build it in a virtual version of the store it is bound for, walk it as a shopper would, then show it to shoppers and watch what they find. The first part tells you whether it can be built. The second tells you whether it works.