The Hardest Categories in Retail Media, and What Actually Makes Them Work

Most retail media case studies come from categories where the shopper researches, compares and buys with real intent.

Electronics and Beauty Categories where a piece of media has room to change a decision.

Data from harder categories, ones built on habit, frequent purchase and price, is rarer, mostly because the results are harder to prove and slower to arrive.

Dairy sits at the hard end of that spectrum.

It is bought frequently, chosen with low consideration, and heavily shaped by in store promotions and prices.

Each of those traits makes it difficult to prove that a specific piece of media, rather than a discount sticker or plain habit, is what drove a given sale.

We recently ran a continuous retail media program for a dairy portfolio in Saudi Arabia to test whether the channel could earn its place in a category like this.

The approach, and what changed as a result, is a useful reference point for any brand in a high frequency, low consideration category weighing up whether retail media is worth the investment.

Why frequency and habit work against attribution

Three things make dairy, and categories like it, harder to attribute than most retail media case studies suggest.

The purchase cycle is short.

A shopper might buy the same product weekly, which leaves little time for a single ad exposure to register as the deciding factor rather than one input among many.

The consideration is low.

Most dairy purchases are close to automatic, built on habit and brand familiarity formed long before the shopper opens a retailer app.

Media has to work against that inertia, not simply add to a research process that barely exists.

The price sensitivity is high.

Trade promotions and price cuts are constant in this category, and a promotion running at the same time as a media campaign makes it genuinely difficult to separate what the discount achieved from what the media achieved.

None of this means retail media cannot work in these categories.

It means the bar for proving it works is higher, and the methodology matters more than in categories where intent is already doing most of the work.

What actually changed the outcome?

Three shifts in approach mattered more than any single piece of creative or targeting.

The retail media team closed the loop against real retailer sales data rather than clicks or impressions, so media exposure was tied directly to what a shopper actually bought.

This is a basic requirement for a category this promotion heavy, since reach and engagement metrics on their own cannot separate media impact from a price cut running in parallel.

Performance was reviewed weekly, not at the end of a campaign.

The budget moved toward what was proven to convert within the flight itself, rather than waiting for a post campaign report to inform the next one.

In a category with this much promotional noise, waiting a full cycle to learn what worked meant learning it too late to act on.

The programme ran as a continuous, always-on effort rather than a single burst.

This mattered because dairy's short purchase cycle and habitual buying pattern take longer than a typical campaign window to shift.

Early months in this kind of category tend to look unremarkable.

The signal shows up later, once the model has had a few purchase cycles to learn from.

What the numbers showed?

One brand in the portfolio, five months into the programme, saw its ROAS climb from 1.8x to 3.7x as the account learned what worked.

A second brand, three months in, held ROAS steady at 2.2x to 2.3x while its volume scaled more than three times over the same period, showing the pattern holds as volume grows, a stronger signal than a strong result on a small scale alone.

Across both brands, the programme generated more than SAR 1 million in attributed revenue, at a blended ROAS of 2.3x, close to the category benchmark of 2.5x and still climbing as the later months of data came in.

What does this mean for brands weighing retail media in hard categories?

A few patterns from this programme are worth carrying into any similar evaluation.

Expect the first month or two to understate the channel's potential.

If a category is genuinely low consideration and promotion heavy, the model needs time and repeat purchase cycles to separate media impact from everything else driving a sale.

Judging retail media in a category like this on a single month's return risks cutting a channel before it has had the chance to learn.

Choose a measurement approach that closes the loop against actual sales, not proxy metrics.

In categories with heavy trade promotion, clicks and impressions cannot tell you whether media or a discount drove the purchase.

Attribution tied to verified retailer sales data is what makes the comparison meaningful.

Review often enough to act within the flight, not only after it.

The teams that get the most out of a programme like this treat it closer to a live operations desk than a campaign that gets a report card at the end.

The takeaway

Hard categories are not a reason to skip retail media.

They are the categories where proving the model actually matters, because the easy wins that intent driven categories offer are not available.

In our experience, a programme that is given the time, the right measurement approach, and a weekly rhythm of decisions tends to earn its place in the plan, even in a category everyone assumed was too difficult to prove.

Ready to Make Retail Media Accountable to Real Sales?

Platformance helps brands connect retail media investment with verified retailer sales, measurable revenue, and ongoing performance optimisation.

See how our retail media solutions help brands turn retailer data into measurable business outcomes.