Comparing store coverage with product distribution

Compare an in-store media footprint with the stores and dates on which shoppers could buy the advertised product. More media stores do not necessarily mean more useful coverage. To compare placement groups fairly, define clear bases for product availability, delivery and outcomes.

For a UK grocery illustration, compare Tesco with Sainsbury’s on the same product set, comparable store formats and flight dates. The comparison should show how much of the product’s distribution had verified media delivery, not declare a winner from chain-wide store or sales totals.

Build comparable groups

Choose groups that answer the decision at hand: entrance versus in-aisle placements, different store formats, or a broad network versus a smaller category-focused group. For the same reporting period, request media locations and active dates, product ranging and stock information, relevant traffic, delivered plays or broadcasts, and sales for the agreed product set.

Distribution changes over time. A period-end ranging count cannot describe an entire flight if products were added or removed during it.

Store count describes footprint, but stores of very different sizes should not automatically receive equal weight in an outcome comparison.

For a Tesco–Sainsbury’s comparison, calculate each placement group separately within each retailer, then compare like with like. Retailer names alone do not establish that the stores, placements or results are comparable.

Request dated ranging and stock records from each retailer, delivery logs from the media operator, and traffic and sales records for the same stores and dates. Agree how each record defines ranging, in-stock status, a visit and a delivered play before calculating rates.

Put the denominator beside every result

Define store-days, unknown ranging status, closures and stock gaps. Category-sales share can be useful for a different question, but its denominator is category sales rather than shopper visits; do not treat the two rates as interchangeable. None of these measures is shopper reach.

A group with higher total sales may simply contain larger or busier stores. A group with a stronger sales rate but few eligible locations may be worth selective use. These are interpretations, not campaign results.

Count eligible store-days only while a store is open and the product is ranged; report closures and unknown ranging status separately. Stocked store-days are the eligible days when the product was available for sale, and delivered store-days are those stocked days with at least one verified play or broadcast.

Report distribution availability as stocked store-days divided by ranged store-days. Report placement coverage as delivered store-days divided by stocked store-days, and end-to-end coverage as delivered store-days divided by ranged store-days; show each numerator, denominator and rate.

Where booked-play counts are available, report verified plays divided by booked plays as a separate delivery measure. For sales outcomes, use sales per stocked store-day or sales per retailer-reported visit for the same stores and dates, and state which denominator is used.

Investigate other differences

Compare store format, local demand, baseline product sales, price, promotions, shelf position, stock and flight dates. An aisle group concentrated in stores already strong in the category could outperform an entrance group without aisle placement causing the difference.

Store-level data, when available, can show whether a small number of high-volume shops dominate the group average. A national total cannot reveal that pattern. Record missing data and small groups before drawing a conclusion.

If visits are the denominator, use the retailer’s count for the same stores and dates, and record how that count is defined. Sales per visit and sales per stocked store-day answer different questions and should not be compared as if they were the same rate.

Decide what the comparison can support

Set a decision rule in advance for acceptable stocked coverage, verified delivery and the outcome measure that would warrant another booking. There is no universal threshold.

For a retailer comparison, set the minimum coverage and delivery levels and the required outcome difference before reviewing results. Prefer one placement group only if it clears both minimums and meets the agreed outcome rule on comparable stores, dates and product availability; otherwise report no preferred group.

An observational comparison can guide placement or identify a question for a test. It cannot isolate incremental sales when stock, traffic and other conditions differ. A causal spending claim needs a suitable test design.

Pros and Cons of Using Sales per Visit vs Sales per Stocked Store-Day in UK In-Store Campaigns