How to measure your shelf share in a store — step by step
Shelf share is estimated by eye in most stores, so it changes between reps and between visits. Here is a practical method that produces a number you can defend to management.
"We hold about a third of the shelf" gets said in a lot of meetings, and nobody knows where it came from. The problem is not that the estimate is always wrong — it is that it is not comparable. You cannot tell whether things improved or got worse, because every visit measured differently.
First: agree on the unit
Before any visit, settle what you are counting in. Mixing units is the single biggest reason numbers disagree:
- Facings: packs visible in the front row. Fastest, and the most practical at scale.
- Centimetres of shelf: more accurate across mixed pack sizes, and far slower to execute.
- Number of shelves: very fast but coarse — a top shelf is not an eye-level shelf.
Second: fix the category boundary before counting
The common failure is one rep measuring "juices" as the main shelf only, while a colleague includes the fridge. Both numbers are correct and cannot be compared.
Define, per category, what counts and what is recorded separately — and put it in the visit checklist itself, not in a WhatsApp broadcast.
Third: separate what cannot be compared
Three cases must always be recorded apart, because merging them hides the problem instead of surfacing it:
- Store fridge versus company fridge: being in the retailer's fridge is not the same as being in one you supplied. You compete for the first; you own the second.
- Ambient shelf: many categories cannot use it at all. Record those as not applicable with a written reason, not as zero — zero reads as failed execution, not applicable reads as a correct decision.
- Secondary display: temporary extra space. Folding it into base share produces a phantom improvement that vanishes when the campaign ends.
Fourth: make a partial visit impossible
The dangerous number is not the wrong one. It is the incomplete one that arrives looking complete — a rep measured eight categories out of twenty-six and submitted, and a campaign decision was built on it.
The fix is operational, not supervisory: make the visit a sequence, category by category, with visible completion, and do not allow it to close until it is done. Then you never have to ask anyone whether they finished.
Fifth: turn it into a weekly indicator
The raw number does not help management. Three things do: average share per chain, the count of stores that fell below an agreed floor, and non-execution reasons categorised.
The third matters most and is skipped most often. The moment you know that forty percent of non-execution is out-of-stock rather than store refusal, the decision changes entirely — from pressing the field team to reviewing supply.