Recalculating reported like-for-like growth on a consistent store cohort routinely moves the figure — sometimes by more than the underlying trading did. Refit timing and the treatment of stores in their thirteenth month explain most of the gap.
What does like-for-like actually measure?
Sales growth from stores open in both comparison periods, excluding openings and closures. It exists to separate trading performance from estate expansion, which is a legitimate and necessary distinction.
The definition of "open in both periods" is set by the company. That single choice does most of the work in the resulting number, and it is disclosed in a way that is technically complete and practically invisible.
The disclosure problem is one of granularity rather than of transparency. Companies do state their convention, usually in a note, and almost none state the composition of the base in a form that would let an outsider re-run the calculation. Knowing that refits are included tells you very little without knowing how many of them there were and when.
How does cohort choice change it?
Materially. When a store enters the comparable base, how refits are treated, and whether temporary closures are excluded can each move the figure by more than the underlying trading did.
None of this requires bad faith. Every convention has a defensible rationale; the point is that the conventions differ between companies and periods, and a like-for-like number is not comparable across either without checking them.
Conventions also change, which is the harder version of the same problem. A company that alters when a store enters the base creates a discontinuity in its own series, and the year of the change is the year the number is least comparable to anything. That is the point at which independent verification is worth most and is least often commissioned.
What can inflate a reported figure?
Refit timing is the most common. A refitted store posts a sharp uplift that decays over subsequent quarters, so a comparable base weighted toward recent refits flatters the current period and drags the next one — which is what a check run before the print is looking for.
Excluding underperforming stores for a definitional reason — a temporary closure, a change of format — has the same effect and is harder to see from outside.
The definitional exclusions are the ones worth asking the trade about, because they are the least visible externally. Store managers and regional staff know which sites were closed briefly, refitted, or reformatted, and that list compared against the reported base is usually enough to establish whether the exclusions were routine or selective.
How do you verify it independently?
Through store and regional managers on footfall, basket and refit timing, and through suppliers on order patterns across the estate. Neither gives you the number; together they tell you whether the direction is real.
Supplier evidence is the stronger of the two for direction, because orders across many stores aggregate away the local noise that dominates any single site — the same reason distributors see a brand's demand before the brand does.
Suppliers are also the practical route where a retailer's estate is large. Speaking to enough store managers to cover a national estate is not feasible inside a diligence window, while a handful of suppliers with national coverage can describe the pattern across it. The trade-off is that suppliers see their own category rather than the whole basket.
Who can tell you the truth?
Regional and area managers know footfall and basket movement across a group of stores, which is the level at which trends become visible. Store-level staff know their own site and generalize badly from it.
Suppliers know whether growth is broad or concentrated, which is the question a like-for-like number cannot answer and which frequently matters more to anyone tracking the category.