Data strategy is crucial to crushing the retail media game

Will Margaritis often feels like he’s flying a plane while building it.

Since he’s working in the rapidly evolving retail media space as IPG Mediabrands’ SVP of global commerce, that sentiment makes sense. After accelerating at the start of the COVID-19 pandemic, retail media has taken off. With the space projected to grow 10.5 percent to a $141.7 billion business in 2024, we are reaching the category’s next stage.

Retail data comes through as a simple concept: Retailers have a ton of first-party purchase data they are willing to share in a privacy-conscious way. Advertising leaders who The Current spoke with agree that data is extremely valuable, but how to make sense of it and unify it is another story. From retailers measuring attribution differently, including offering different lookback windows, to the challenges of tracking each piece of advertising back to a sale across multiple channels and matching customer profiles, the retail data landscape offers a lot of complexity.

Margaritis says unifying measurement is either the biggest challenge or the biggest opportunity in the retail data space.

“It just creates a lot of noise and can be difficult to get to an apples-to-apples comparison,” he tells The Current. “[We’re] ultimately attempting to unify [that data] and really have the right data structures and the right algorithms around it to pick out all of the critical insights from that noise.”

Ninety-one percent of advertisers in the U.S. are planning to maintain or increase their investment in retail data in the next few years, according to a recent survey  from The Trade Desk Intelligence and Material+. On top of that, 68 percent of advertisers surveyed that don’t currently have a retail media strategy are reportedly planning to create one in the future, meaning simplifying the complexity around retail data is paramount.

Marketers use this data to power campaigns on retailers’ sites, using the freshness of the real-time purchase behavior that wasn’t previously as accurate or as available. By targeting the customers most likely to be interested in buying their products, brands can find the most relevant consumers and reduce wasted ad spend.

Increasingly, retailers are also making this data available for ad campaigns off-site, which is emerging as an innovative way to leverage a retailer’s customer data for modeling and targeting outside of retailers’ properties. Margaritis believes off-site media’s potential is unlimited, as channels like connected TV (CTV) and digital out-of-home (DOOH) allow performance and brand awareness to strengthen one another side by side.

Connecting data for the full picture

Overall, quality and scale are the two characteristics that separate the best from the rest when it comes to retail data, according to Amperity co-founder and chief technology officer Derek Slager. Amperity works with brands to clean up their customer data sets.

Insights from Amperity shared with The Current show 46 percent of all its clients’ customer profiles that were originally marked as high value were actually wrongly identified. This mistake comes from a lack of quality data, according to Slager, and includes single consumers marked several times in databases or in silos disconnected from other sources.

Identifying that data “makes a massive difference,” Slager tells The Current. “Even in cases where maybe you’re only de-duplicating 5 or 6 percent, it doesn’t sound very impactful, but given the outsized value of that in your highest-value customers, even small improvements in building that data foundation can have an enormous impact. And that scales out directly to retail media.”

Margaritis sees another side of the data debate. He wants all the data he can get his hands on so that his team can parse it all out to find the gold. Even with the scale of the data they’re getting, it’s brought down to the quality insights.

To emphasize this point, Margaritis uses the example of selling different-size bags of pretzels. A pack that’s 15 ounces versus 20 ounces is relatively the same product, but each is counted as a different product. And if an agency like IPG Mediabrands is running ads for a pretzel client with one retailer but the consumer ends up buying a smaller pack through a different retailer, the agency has to do the legwork to attribute that sale.

Tracking attribution across multiple retail media networks would unlock a treasure trove of insights for marketers and foster smarter planning.

“Sometimes you do need that mountain to truly find the right insights,” Margaritis says. “It might take longer to get through [it], but the more data you have, the more you can learn from, the more you can see how different media channels or products are impacting each other, which becomes critical for efficient planning.”

The Current is owned and operated by The Trade Desk Inc.