When Amazon unveiled new digital ad products in 2019 to reach people who shop at its few brick-and-mortar AmazonFresh stores, company watchers forecasted the inevitable: eventually, advertisers would be able to connect in-store purchases made at Amazon-owned Whole Foods back to the digital ads Amazon showed them.This February, the company did just that. For the first time, Amazon provided advertisers with attribution data that showed when ads bought through its demand-side platform led to purchases made in-person at its more than 500 Whole Foods Markets across the U.S. Although there are gaps in the data and challenges in using it, advertisers already say they’re seeing new data nuggets illuminating more about how people choose what to buy.
“There are instances where we see a product where maybe it’s not [a brand’s] best-selling product on Amazon.com, but then we actually see that lesser-best-selling product being purchased within Whole Foods,” said Maddie Kaseeska, associate director of marketplace programmatic at Tinuiti, which handles ad campaigns for clients in platforms including Amazon. The agency is “actively testing” the new attribution feature with consumer-packaged-goods advertisers such as paper goods producers and a protein bar brand Kaseeska declined to name.
The data has revealed that a top-selling flavor of protein bars made by that brand purchased in physical stores, for instance, is different from the most-bought flavor online, she said. Such information might inform future digital ad targeting based on specific flavors people bought online as opposed to in actual stores, where shoppers aren’t necessarily influenced by online reviews or star-studded product ratings.
“If we are running our display ads on Amazon.com or off-site via the DSP, we’re leveraging that first-party data [and] able to create that control group in terms of who we’re targeting,” said Kaseeska.
In general, agency execs testing the service for advertisers say Amazon’s Whole Foods ad attribution is another step along the company’s anticipated path toward connecting digital to in-store data points and revealing how advertising affects actual in-store purchases. “Now we’re able to tell a more holistic story and narrative in terms of how our display ads are impacting the consumer off of Amazon as well,” said Kaseeska.
A new front in Amazon’s ad war with Walmart
Linking online ads to in-person purchases could help give the e-commerce giant an edge as it competes with the likes of Walmart, which operates more than 11,000 stores worldwide.“The way we look at it is it’s the first of many steps Amazon is going to take in connecting physical properties to digital,” said Nich Seo, director, go-to-market for commerce at data-centric consultancy MightyHive. While he stressed that Amazon’s decision to add the new attribution reporting was not necessarily a direct swipe at Walmart, he said, “The move toward physical attribution is meant to be competitive with Walmart.”
Walmart is working with The Trade Desk to build a demand-side platform that is expected to launch in the second half of this year. So far, media buyers say they haven’t heard much about when they’ll be able to test the Walmart DSP. Both Walmart and Kroger — another Amazon competitor in the grocery sector — offer attribution data to advertisers that help show the impact of digital ads on in-store and digital purchases.
However, Amazon and Kroger have strengths where Walmart does not, and vice versa, said Ryan Flanagan, CEO of Nuanced Media, a full-service e-commerce and Amazon marketing agency. “The primary difference is the scale of the marketplaces, particularly when it comes to e-commerce compared to brick and mortar,” said Flanagan.
This year Amazon is expected to generate $367.2 billion in U.S. e-commerce sales, according to eMarketer. That eclipses the projected sales of second-ranked Walmart, which is expected to bring in $64.6 billion, as well as Kroger at $15.0 billion. However, e-commerce remains a fraction of the overall retail marketplace. This year e-commerce will account for 15.5% of the $5.9 trillion in total U.S. retail sales per the research firm’s forecast.
“Walmart’s and Kroger’s brick and mortar footprints are significantly larger than Amazon’s,” Flanagan said. “When you are speaking about true omnichannel marketing, all of these companies have a significantly different footprint, and it will be interesting to see how this all shakes out as the importance of online sales becomes more important than in-store sales due to purchasing trends moving away from in-store purchases.”
Though it is unclear whether the new Amazon reporting could lead advertisers to shift budget away from other retail media, it could lead to incremental budget and spending on Amazon digital ads if brands see a positive impact on awareness metrics as tracked through purchase data over time, said Kaseeska.
The Whole Foods reporting won’t show the whole picture, though. It only connects the digital ad dots to in-store purchases if people are Amazon Prime members and then only when they scan their Prime membership account codes at checkout. “There are a lot of steps with that to make sure that works,” Flanagan said.
Amazon reported in its recent earnings report that advertising revenue within North America and international segments accelerated during the first quarter of 2021. Amazon declined to comment on the record for this story.
Location, location, location
Whole Foods’ in-store attribution tracking by city or state allows companies to test new products in two or three markets before rolling them out across the country, Flanagan said. Plus, if Amazon is evaluating a new brand’s viability in a test market, it gives the brand an incentive to run geographically-targeted digital campaigns through Amazon’s DSP in conjunction with that product pilot, he said.
Ultimately, said Flanagan, without insight into purchases made in stores, advertisers don’t know how much sales lift they can attribute to their marketing. He said the attribution reporting also can help brands determine the lifetime value of specific customers, allowing them to use the data “to test a hypothesis and build out real metrics and [key performance indicators].”
Still, in addition to gaps in the data, if people who were shown ads aren’t Prime members or don’t scan their account information when they make purchases, Seo said there are other hurdles to making the new reporting relevant. Right now, because there’s nothing to compare the Whole Foods in-store attribution data to in an apples-to-apples way, he said, “It’s hard to benchmark.”