Retail Media Advertising: What the Numbers Don’t Tell You

Retail media advertising is the practice of brands buying ad placements within a retailer’s owned digital environment, on-site search, sponsored product listings, display banners, and increasingly off-site inventory that a retailer packages using its first-party shopper data. It has grown fast because it sits close to the point of purchase and because retailers, sitting on rich transaction data, can make a compelling measurement story. Whether that story holds up under scrutiny is a different question.

The category is now large enough that most major consumer brands have a retail media line in their media plan. The question worth asking is not whether to be in it, but how to be in it without letting the channel’s built-in attribution logic do your thinking for you.

Why Retail Media Grew So Fast

The timing was not accidental. Third-party cookies started their long, slow collapse. Apple’s App Tracking Transparency reshaped mobile attribution. Brands that had built performance models on cheap, trackable digital inventory suddenly found themselves paying more for less signal. Retailers stepped into that gap with something that looked like a solution: first-party data tied directly to purchase behaviour, sold back to brands as a media product.

Amazon built the template. Walmart, Kroger, Target, Boots, Tesco, and dozens of others followed. The pitch is structurally compelling. You advertise inside an environment where people are already shopping. The retailer can tell you whether someone who saw your ad then bought your product. The loop looks closed.

I spent a significant part of my career running agency P&Ls and managing large media budgets across multiple categories. One thing I learned early, and had to relearn a few times, is that channels which generate clean-looking numbers attract budget faster than channels that generate honest ones. Retail media has that problem baked in.

That does not make it a bad channel. It makes it a channel that requires more rigour than most marketing teams apply to it.

What Retail Media Actually Includes

The term covers more ground than most people realise when they first encounter it. The core formats are sponsored product listings (appearing in search results on the retailer’s site or app), sponsored brand placements (banner-style units that appear in category pages or homepages), and display advertising served within the retailer’s digital estate. These are the formats most brands start with because they are closest to the shelf and easiest to connect to sales data.

Beyond that, most major retail media networks now offer off-site inventory. This means the retailer uses its audience data to serve ads on external platforms, social media, programmatic display, connected TV, and in some cases audio. The retailer’s data informs the targeting, but the impression is served outside the retailer’s own environment. This is where the proposition gets more interesting and where the measurement claims get shakier.

There is also in-store media, digital screens, audio at the point of sale, and sampling activations that some retailers now bundle into their media networks. This is genuinely new territory and worth watching, though the measurement frameworks for in-store are still being worked out by most networks.

If you are thinking about how retail media fits into a broader go-to-market plan, the Go-To-Market and Growth Strategy hub covers the full planning architecture, from channel selection to commercial objective setting, and gives retail media the context it needs to be evaluated properly rather than in isolation.

The Attribution Problem Nobody Wants to Say Out Loud

Earlier in my career I overvalued lower-funnel performance metrics. I was not alone in this. The whole industry was moving in that direction, chasing trackable conversions and crediting the last touchpoint that could prove a click. What I came to understand, slowly and through enough budget cycles to see the pattern, is that a meaningful portion of what performance channels claim credit for was going to happen regardless.

Retail media has the same structural issue, amplified. When someone searches for your brand name on a retailer’s site and then clicks a sponsored listing for your product, did the ad cause the purchase? Almost certainly not. The person was already looking for you. You paid for a placement to intercept intent that existed before the ad was served. The retailer’s dashboard will show a conversion. The ROAS will look strong. None of that tells you whether the spend was necessary.

This is not a fringe concern. It is the central measurement challenge in retail media, and it is one that retailer-provided reporting has a structural incentive to underplay. The retailer profits when you spend more. Their attribution model is not built to show you when your spend was redundant.

The honest version of retail media measurement requires incrementality testing: running controlled experiments where some exposed audiences are compared to unexposed control groups, then measuring the difference in purchase behaviour. A small number of sophisticated brands are doing this. Most are not, because it requires pushing back on the retailer’s preferred reporting methodology and accepting that the numbers will look less impressive than the dashboard suggests.

This pattern is well-documented in broader performance marketing literature. Go-to-market execution is getting harder across the board, partly because the easy attribution wins of the last decade masked how much growth was being borrowed from existing demand rather than created through marketing.

Where Retail Media Genuinely Earns Its Place

None of the above means you should pull budget from retail media. It means you should be clear about what it is actually doing for you.

Sponsored search on retail platforms is a legitimate defensive tool. If your brand is well-known and your category is competitive, letting a competitor’s sponsored listing appear above your organic result when someone searches your brand name is a real commercial problem. Paying to defend that position makes sense. It is not growth marketing. It is shelf maintenance in a digital environment. Treat it as such, budget it accordingly, and do not expect it to generate new buyers.

Category-level sponsored listings are more interesting. When someone searches “protein bars” or “moisturiser SPF 50” and your product appears, there is a reasonable case that the placement influenced a consideration that was not fully formed. That is closer to genuine demand capture from a competitive pool, and it is worth investing in if your product is genuinely competitive on the shelf.

The retailer’s first-party data used for off-site targeting is where I see the most underutilised potential. A retailer that can tell you which of its shoppers buy in your category but have not bought your brand specifically, and then let you target that audience off-site with brand-building creative, is offering something that was genuinely hard to do before retail media networks existed. That is audience development, not just intent capture. It connects to the broader principle that growth requires reaching people who are not yet customers, not just converting people who already are.

Think of it this way. Someone browsing a category page on a retailer’s site, looking at competitors, is not yet your customer. They are in the market. Getting in front of them, with the right message, at that moment, is meaningful. That is where retail media earns genuine credit rather than borrowed credit.

How to Structure a Retail Media Investment That Makes Commercial Sense

The first thing to get clear on is what you are trying to do. Retail media can serve multiple objectives, but it cannot serve all of them equally, and the format mix you choose should follow the objective, not the retailer’s rate card.

If the objective is protecting existing market share in a competitive category, sponsored brand search is the right tool. Budget it as a cost of doing business, measure it against share of search rather than ROAS, and do not expect it to grow your customer base.

If the objective is winning new buyers who are already in-market in your category, category-level sponsored listings and category page display are the right formats. Measure them against new-to-brand purchase rate, which some retailers now provide, rather than total sales. If the retailer cannot give you new-to-brand data, that is a meaningful gap in the measurement framework and worth negotiating before you commit significant budget.

If the objective is building awareness among category buyers who have not yet considered your brand, off-site inventory using the retailer’s audience data is the relevant format. This should be measured against reach and frequency metrics among the target audience segment, with purchase behaviour tracked over a longer attribution window than the standard 14 or 30 days most retail media networks default to.

Running agencies at scale taught me that the biggest budget waste rarely comes from choosing the wrong channel. It comes from using the right channel for the wrong objective and then measuring it in a way that confirms whatever you hoped to see. Retail media is particularly vulnerable to this because the measurement infrastructure is controlled by the party selling you the media.

A useful framework from the broader growth strategy literature is thinking about market penetration as the primary growth lever for most established brands. Retail media, used well, can support penetration by getting your product in front of category buyers who have not chosen you yet. Used poorly, it recycles spend on buyers who were already going to buy.

The Retailer Relationship and What It Actually Gives You

There is a commercial dynamic in retail media that does not get discussed enough. When you buy media through a retailer’s network, you are not just buying advertising. You are also deepening a commercial relationship with a trading partner who has significant leverage over your distribution, shelf placement, and promotional terms.

This is not inherently a problem. A strong retailer relationship has genuine value. But it does mean that retail media investment decisions sometimes get made for relationship reasons rather than pure media efficiency reasons, and those two things should at least be acknowledged separately in the planning process.

The brands that handle this best are the ones that have a clear internal view of what they expect retail media to deliver on purely commercial terms, and a separate view of what value they place on the relationship dimension. Conflating the two leads to media plans that are hard to evaluate honestly.

It also leads to a specific failure mode I have seen more than once: brands increasing retail media spend because a key account manager signals it would be well received, then presenting the resulting ROAS numbers internally as evidence of media efficiency. The spend may have been justified on relationship grounds. It was not justified on the grounds being used to defend it.

Honest commercial planning requires separating those conversations. What does the media buy deliver on its own terms? What is the relationship value worth? Both are legitimate inputs. They should not be mixed into a single ROAS figure and presented as one thing.

Building Your Own Measurement View

The practical implication of everything above is that you need a measurement framework that is not entirely dependent on the retailer’s reporting infrastructure. This does not have to be complicated. It does have to be independent.

At minimum, track total brand sales within the retailer over time and look for correlation between media investment periods and sales trajectory. This is not proof of causation, but it gives you a directional view that is not filtered through the retailer’s attribution model.

Where possible, run hold-out tests. Pause retail media in one region or one product category for a defined period and compare performance against a matched control. Most brands resist this because they are nervous about what the results will show. That nervousness is itself informative.

Track new-to-brand metrics wherever the retailer provides them. New-to-brand purchase rate is a more honest indicator of whether retail media is growing your customer base than total ROAS. If a retailer cannot or will not provide new-to-brand data, that should factor into how much weight you give their reporting.

Use market mix modelling if your budget justifies it. Retail media is now a standard input in most MMM frameworks, and it gives you an independent view of contribution that sits outside the retailer’s closed loop. BCG’s work on go-to-market strategy consistently points to the value of having a commercial model that integrates across channels rather than evaluating each in isolation.

The brands that are most commercially sophisticated about retail media are the ones that treat it as one input into a full-funnel model, not as a self-contained performance channel with its own measurement logic. That distinction matters more as retail media budgets grow and the channel takes a larger share of total investment.

What the Next Phase of Retail Media Looks Like

The networks are maturing fast. The early phase was dominated by on-site sponsored listings with basic keyword targeting. The next phase, which is already underway at the larger networks, involves more sophisticated audience segmentation, off-site and connected TV inventory, in-store digital media, and increasingly, collaborative data arrangements where brands bring their own first-party data to enrich retailer targeting.

Creator-led content is also entering the retail media conversation. Some networks are beginning to integrate creator placements into their off-site inventory, which creates interesting possibilities for brands that want to combine the retailer’s audience data with content that does not look like a traditional ad. Integrating creators into go-to-market campaigns is becoming a more structured discipline, and retail media is one of the environments where that integration is starting to show up.

The standardisation question is also moving. One of the persistent frustrations with retail media is that every network has its own reporting interface, its own attribution methodology, and its own data taxonomy. Brands running campaigns across multiple retailers are essentially comparing apples to oranges in their reporting. Industry bodies are working on common standards, but the pace is slow because standardisation reduces the information asymmetry that benefits retailers in commercial negotiations.

The brands that will be best positioned as the category matures are the ones building internal capability now: people who understand the channel mechanics, who can interrogate retailer reporting rather than accept it, and who can connect retail media investment to broader commercial outcomes rather than treating it as a separate budget line managed by the shopper marketing team.

Retail media is not going away. The data asset that retailers hold is real, the proximity to purchase is real, and the scale of the major networks is real. What is not real is the idea that you can hand over budget, accept the dashboard at face value, and call it a growth strategy. Growth requires reaching people who are not yet yours. Retail media can do that. Most of the time, it does not, because the path of least resistance in the channel runs straight to the people who were already going to buy.

If you want to think about retail media in the context of a broader commercial plan, the Go-To-Market and Growth Strategy hub is the right place to start. The channel decisions are downstream of the strategic ones, and retail media is no exception to that rule.

About the Author

Keith Lacy is a marketing strategist and former agency CEO with 20+ years of experience across agency leadership, performance marketing, and commercial strategy. He writes The Marketing Juice to cut through the noise and share what works.

Frequently Asked Questions