Every retailer cannot become its own operating system

One of the biggest challenges in retail media is fragmentation.

Large retailers are building networks with their own naming conventions, reporting structures, audience logic, buying workflows, and data policies. Smaller retailers are deciding whether to build, partner, or participate through aggregators. Brands and agencies are trying to compare opportunities that may look similar in a presentation but behave very differently in practice.

This creates a familiar ad tech problem: growth without interoperability.

A marketer may want to compare reach, frequency, return on ad spend, incrementality, or category performance across retail media partners. But metrics are not always defined consistently. Attribution windows differ. Audience construction varies. Available inventory changes by network. Some placements are fully on-site, others are off-site, others sit somewhere between retail media and broader commerce media.

The result is that brands face more opportunity, but also more operational tax.

If retail media wants to keep attracting budgets beyond the largest endemic advertisers, it needs to become easier to understand and easier to evaluate.

Programmatic rails can help — but only if standards follow

Programmatic infrastructure offers an obvious path to scale retail media beyond retailer-owned pages. It can bring auction mechanics, audience activation, deal-based buying, and access to diverse supply. It can help advertisers activate retailer data across a broader range of inventory, and it can help retail media networks reach buyers already working through DSP-led workflows.

But programmatic cannot simply absorb retail media as-is without friction.

Retail data is sensitive. Audience quality depends on recency, transaction context, and taxonomy. Retail media often carries stricter commercial expectations around closed-loop measurement. And not every retailer wants to expose its media business to a fully commoditized environment.

The winning model will not be “turn all retail media into open exchange buying.” It will likely be a combination of:

  • on-site retail placements with improved automation
  • off-site programmatic activation under clear data rules
  • curated commerce audiences
  • private marketplace structures for premium supply
  • more standardized measurement and reporting frameworks

In other words, retail media and programmatic are moving closer together, but they will need to meet in the middle.

Data alone does not make media valuable

Retail media’s strongest asset is often framed as first-party purchase data. That is fair — but incomplete.

Data is only useful when it can be applied responsibly and effectively. A retailer may know that a household buys pet food, premium coffee, or consumer electronics. The value of that knowledge depends on whether it can be activated at the right moment, with the right consent structure, in the right media environment, and measured in a way the advertiser trusts.

Poorly governed data does not become premium simply because it comes from a retailer. Weak inventory does not become high-value simply because a commerce segment sits on top of it. And a campaign should not be considered successful merely because it reports sales that might have happened anyway.

Retail media’s next stage will depend less on possessing data and more on proving that data improves media outcomes.

The measurement conversation will become tougher

Retail media has benefited from a compelling promise: it can connect advertising to transactions. That promise is powerful, especially compared with channels where outcome measurement is indirect.

But as retail media expands into upper-funnel objectives and off-site environments, measurement becomes more complicated. Last-touch attribution can make performance look cleaner than it really is. Retailer- specific reporting can be hard to compare. Closed-loop data may still leave open questions about incrementality, halo effects, and media’s contribution across the full path to purchase.

Sophisticated advertisers will increasingly ask:

  • Did this campaign reach a genuinely incremental shopper?
  • Did it shift brand preference or only capture existing demand?
  • How should retail media be compared with search, social, CTV, and standard programmatic?
  • Are we paying for insight, access, or both?

The networks that answer these questions clearly will build longer-term trust.

There is a role for the open ecosystem

Retail media is often described as another wave of walled gardens. That may be true for some large platforms, but it should not be the only path.

There is an opportunity for open ad tech to support commerce media in a more interoperable way. SSPs, DSPs, data platforms, and clean-room environments can help connect high-quality supply with retailer signals without collapsing everything into closed infrastructure.

For publishers, this may create new monetization opportunities through commerce-informed demand. For buyers, it may provide more flexibility in how retail audiences are reached. For retailers, it may extend the value of their data without requiring them to reinvent every layer of the ad stack.

The condition is transparency. Buyers need to understand what data is being used, how it is segmented, where it is activated, and what measurement means. Sellers need to know how their inventory is packaged. Consumers need their data choices respected.

Retail media is becoming real media

The most important shift may be conceptual. Retail media is no longer a side budget owned only by shopper marketing teams. It is becoming part of broader media planning. That will expose it to higher expectations.

It will need stronger standards. Better interoperability. More credible measurement. Clearer data governance. And, in many cases, a more mature relationship with programmatic infrastructure.

That is not a threat to retail media. It is a sign that the category is growing up.

The next phase will belong to companies that make commerce signals useful at scale without turning the advertising ecosystem into an even more fragmented maze.