How category performance is measured in retail media
Tracking splits into two layers. Outcome metrics, meaning sales, units sold, ROAS, and advertising cost of sales (ACOS), show what your ad spend returned. Units sold add the volume check a strong ROAS can hide. Visibility metrics, meaning impressions, click-through rate (CTR), and share of voice, show how much of the category's attention you captured. Amazon's category benchmarking tracks impressions, CTR, ROAS, and ACOS for Sponsored Brands advertisers.
The defining habit is relative reading: a CTR or ROAS figure means little until you know where the median sits, so numbers get read against peers. Because each retail media network long reported results its own way, IAB and MRC finalized retail media measurement guidelines in January 2024 to standardize how networks track and report.
Category performance in RMN reporting
Teams meet category performance in network dashboards and downloadable reports. The clearest example is Amazon's category benchmark report (CBR), launched January 28, 2021, for Sponsored Brands campaigns. Vendors and sellers download it from the ad console in North America, South America, Europe, the Middle East, and Asia Pacific. For each KPI (impressions, CTR, ROAS, and ACOS), the report shows your value against your closest peers at the median, the bottom-25% quartile, and the top-25% quartile, over a 90-day lookback window.
The value is in the pairings. Median-level ROAS with bottom-quartile impressions says your ads convert fine but you are spending too little there. Amazon names this use case: identifying "insufficient spend in categories with lower relative impressions". Bottom-quartile CTR with healthy impressions points at the ad itself; the budget is fine. Used with an Amazon keyword research tool, the benchmark helps locate the search terms where that extra spend should land.
Share of voice and share of shelf
Two more metrics track visibility. Share of voice (SOV) is the share of a category's total ad visibility your brand captures: your ad impressions divided by total market ad impressions, times 100. SOV counts paid placements only. Share of shelf (SOS) widens the lens to paid plus organic presence combined: your total footprint on the digital shelf.
Category structure differs sharply by network. Pacvue's analysis found Amazon consumer electronics highly fragmented, with 84 brands competing for top-ten keyword placements and no brand above 10% SOV, while one Walmart category it observed had a single brand holding over 90%. The same SOV number can mean dominance in one group and parity in another, so read it against the network's structure first.
Metrics that matter: units, ROAS, and incrementality
ROAS has long been the gold standard of retail media reporting, and it remains the headline KPI networks benchmark. Its weakness is attribution. Network-reported ROAS can credit ads for sales that would have happened anyway, which inflates the result.
The stricter test is incrementality: sales lift against the category's organic baseline. Retailers argue they are well placed to run it. Kroger Precision Marketing says retailers "know exactly what the baseline organic sales should be in any category we sell," which lets networks report incremental sales alongside attributed ROAS.
Per eMarketer, reported by Digiday in January 2025, retail media should reach nearly 25% of total media ad spend by 2028. Spend that large invites scrutiny of the numbers behind it.
Why it matters for brands and sellers
Category context separates two problems that look the same in a campaign dashboard. Weak ROAS while your impressions sit in the top quartile points to creative, targeting, or pricing work. Weak ROAS on bottom-quartile impressions points to underinvestment, because the ads already convert. Same symptom, opposite fixes, and only the category benchmark tells you which one you have. That difference decides where money goes next.
Category fluency also wins budget arguments. As scrutiny of network-reported metrics grows, brands that frame results against the category baseline as well as their own numbers defend retail media budgets more credibly. Sellers deciding where to compete can start with our retail media solutions for sellers and the retail media database to compare networks side by side.
Related terms
- Retail media network (RMN): where category performance is reported.
- Share of voice: paid share of category ad impressions.
- Incrementality: the lift-versus-baseline test for ad spend.
- Closed-loop attribution: how RMNs tie spend to sales.
- Average order value (AOV): a companion outcome metric.
Frequently asked questions
What does category performance mean in retail media?
Category performance describes how a brand's products and ads perform within a retailer's product category, judged against the other brands in it. Retail media networks track it through ad KPIs such as impressions, click-through rate, ROAS, and ACOS.
How do you measure category performance?
Combine outcome and visibility metrics, then read both against category peers. Amazon, for example, benchmarks impressions, CTR, ROAS, and ACOS against the category median and quartiles. On the visibility side, share of voice divides your ad impressions by the category total.
What is a category benchmark report?
A category benchmark report is a downloadable Amazon Ads report, launched January 28, 2021, for Sponsored Brands advertisers. It compares impressions, CTR, ROAS, and ACOS with your closest peers at the median, bottom 25%, and top 25%, across a 90-day window.
What is the difference between share of voice and share of shelf?
Share of voice (SOV) counts paid ads only: the share of a category's ad impressions your sponsored placements capture. Share of shelf (SOS) adds organic to paid: your total footprint on the digital shelf. Strong SOV can pair with weak SOS.
Is ROAS enough to judge category performance?
No. Network-attributed ROAS can credit ads for sales that would have happened anyway, which flatters category results. The stricter test is incrementality: lift against the category's organic baseline, which retailers like Kroger say they can model for every category they sell.
