Any business with logged-in users and purchase data can announce a media network. Far fewer have what it takes to sustain one.
Retail media began as a straightforward proposition. A retailer holds purchase data and owns a shopping surface where buying intent is unusually concentrated, so it sells advertising against both. The economics worked well enough that the model has spread far outside retail, into travel, financial services, delivery platforms, telecommunications and healthcare-adjacent businesses.
The logic of the extension is sound. Any business with logged-in users, first-party transaction data and reachable inventory can in principle build a network. The gap between principle and practice is where advertiser budget gets lost.
Four conditions for viability
Scale of addressable audience comes first, and it is the one most commonly overstated. What matters is not total customers but the number reachable through the network with sufficient frequency to influence a decision. A large customer base that logs in twice a year is not media inventory in any meaningful sense.
Purchase or intent data specific enough to target on comes second. Transaction data is valuable when it maps to a category an advertiser wants to reach. Data showing that a customer bought something, without category-level detail, supports far less targeting than the pitch implies.
Inventory that a person actually sees comes third. On-site placements on a surface with genuine dwell time behave very differently from off-site extensions where the network is essentially reselling audience segments into someone else's inventory. Both can work. They are not the same product and should not carry the same price.
Measurement the advertiser can verify comes fourth, and it is where the category is weakest. A network measuring its own effectiveness, using its own definitions, with no independent verification, is asking for trust it has not yet earned.
The comparability problem
The most difficult practical issue for a buyer is that networks define their metrics differently. Attribution windows vary. Some count view-through conversions, some do not. What qualifies as an impression differs by placement type. Two networks can each report strong returns on the same budget and the numbers will not be comparable in any rigorous sense.
This is not necessarily deliberate. The category grew quickly and standards have lagged, which is what happens in most young advertising markets. It does mean that any cross-network comparison built from vendor-reported figures should be treated as directional at best.
What to ask before committing budget
How many people can you actually reach in my category, at what frequency, in the regions I care about? What is the attribution window and does it include view-through? Is this on-site inventory or an off-site extension, and how is each priced? What independent verification is available? And what does a like-for-like incrementality test look like, if we run one?
That last question is the useful one. A network confident in its performance will support an incrementality test. Reluctance to be measured independently tells you something the case studies will not.
For advertisers: treat a new network as an experiment with a defined budget and a measurement plan agreed in advance, not as a channel allocation. The category is young enough that this is simply prudent.
How we work. This article was researched and written by the Marketing Hub Media editorial team. We do not republish press releases. Where we cite data we name the source and the method. Corrections are made openly on the article - if you believe something here is wrong, write to info@marketinghubmedia.com.
Filed under Advertising & Media · Get the weekly brief