The key to optimizing AI to drive real retail media outcomes

Retail media brings the promise for advertisers to reach real people relevantly, drive loyalty from existing customers and convert prospects into new customers. It also brings the promise for advertisers to measure the true effectiveness of their advertising—no more “best guesses.”

But with the deprecation of third-party cookies, the rapid advancement of artificial intelligence (AI) and the abundance of mediocre identity solutions out there, now is the time for retailers and advertisers to reevaluate their approach to retail media.

Retailers need to leverage the most enhanced capabilities to optimize their retail media offerings and drive better (and real) outcomes for advertisers. While it probably surprises no one that AI is the key here, it’s crucial that this technology is used correctly to achieve the key to modern retail media success: people-based intelligence.

Let’s look at what exactly this people-based intelligence is, what it’s not and how advertisers and retailers can include it in their retail media strategies.

What is people-based intelligence?

People-based intelligence is two-fold—combining honest identity resolution with next-generation AI intelligence.

Honest shopper identity resolution: What is “honest” in the context of identity? It comes down to retailers giving their advertising partners:

  • the ability to reach all of their shoppers,
  • assurance that they are reaching real people (not fragments of identities) and
  • a guarantee to reach these real shoppers wherever they are.

Many retail media solutions promise these things, but few can deliver.

Most retail media platforms leverage DSPs that evaluate sites first, not people. This eliminates the opportunity to identify and reach 100% of the retailer’s verified shoppers. Evaluating the actual person, whether they are on the retailer’s site or on the open web, ensures all shoppers can be considered and, if your AI solution determines appropriate, engaged.

When these retail media platforms get around to attempting to reach shoppers, they are actually just targeting “audiences of digital identifiers,” which translates to fragments of a whole shopper identity. This could look like messaging three different device IDs that are tied to one individual. The problems here include a large amount of wasted ad spend compounded by dissatisfied consumers annoyed by ad repetition. Studies have not only demonstrated that negative associations for a brand can spike when ads are overserved to consumers, but also that purchase intent can be eroded by ad overexposure.

The industry is also collectively facing third-party identifier deprecation, which means marketers need identity resolution centered around robust first-party data. Retail media solutions rooted in people-based identity anchored to deterministic data elements make it not only reliable in finding the right consumers, but also stable and future-proofed against regulatory and industry shifts. For retail media strategies to function optimally, retailers need to select partners with these identity capabilities; both retailers and the partners with which they work need to be grounded in person-based identity.

Intelligent AI decisioning: Typical retail media offerings use ad servers that, at most, can consider one variable in the milliseconds available for decisioning. But in today’s world, ad servers must be capable of making hundreds of decisions in the same time frame in order for both retailers and advertisers to realize the true promise of retail media.

When AI is applied to the ad server and shopper engagements on the retailer’s owned properties and across the open web, a multitude of variables can be evaluated in the milliseconds before a bid is placed. AI can also learn from each shopper engagement, adjusting the next in real time to improve the outcome. This increase in power and speed enables advertisers and retailers to go far beyond delivering the proverbial right message to the right person at the right time. Now, on top of delivering true relevance to shoppers, they can also determine where to engage these shoppers, how often to engage them and how much to spend to acquire any given impression, or the shopper in total. These advances can make a difference in driving ROAS, creating loyalty and converting prospects into new customers.

Of course, AI is only as good as the data that fuels it. AI needs the right data to train it, and its outputs are only as good as the data inputs it receives. To that end, it’s important to ensure that the shopper data fueling your retail media’s AI comprises known individuals, not “best guesses” or general audiences. Honest shopper identity resolution powers AI to create truly impactful people-based intelligence.

Driving real outcomes

People-based intelligence drives reliable outcomes for advertisers while increasing returns and shopper loyalty for retailers.

Ulta Beauty’s retail media network, UB Media, uses people-based intelligence to help its advertisers/partners identify which potential or existing in-market customers are most likely to purchase from Ulta Beauty, and then connect with them across channels—onsite and offsite—with personalized, relevant advertising.

According to Ulta Beauty CMO, Michelle Crossan-Matos, “A lot of brands put media on the air and aren’t able to directly link it to revenue. Because our retail media network has closed-loop reporting, we can see the direct link to revenue, and that’s pretty powerful.” This is the case as UB Media’s campaign results consistently exceed industry benchmarks across offerings.

Along the same lines, Walgreens Advertising Group (wag) leverages people-based intelligence to ensure real-time optimization. Each brand campaign is continuously optimized in real time across Walgreens’ owned website and the open web, based on verified in-store and online purchase data, to ensure delivery is prioritized to the best prospects.

When describing wag, Tracey D. Brown, president of Walgreens retail and chief customer officer said, “Our ability to take the data and connect at an individual level and then serve up the things that are most important to meet the customer’s needs—that is golden.”

A single shopper view, applied intelligently

When you combine true one-to-one customer identity resolution with next-generation predictive AI capabilities, you arrive at people-based intelligence that levels up retail media campaigns. Of course, none of this will work if advertisers don’t have the freedom to decide where, when, what and how to engage with shoppers.

A complete retail media solution that also is flexible enough to augment retailers’ existing retail media offerings applies people-based intelligence wherever it is needed—always linking via a single view of the shopper across any channel, on the retailer’s or across the open web, applied intelligently in the retail media context.

This article was originally published on Adweek, April 2024.