The CFO Loves Those Retail Media #s. He'd Like to See Them Everywhere.
But should every media employ shopping data?
This post is part of a special three-part series on the state of the video advertising sector, produced in partnership with National CineMedia
The past few years in advertising have been marked by Retail Media Fever.
Now, the entire industry may be hit with a case of the Trump Tariff Flu.
So if you thought that this red-hot ad category was already of great interest - just wait. The worry now is whether brands, led by their ROI-hungry Chief Financial Officers, may go overboard with RM spending, even overdosing on RM data.
“It’s a huge moment in the industry right now,” said Ali Miller, VP of ads product at Instacart. “There have already been so many changes in the world, particularly among CPG brands. That’s been a big sea change already. Now on top of this you’re talking about uncertainty.”
“Brands do turn to more measurable media [during downturns],” she added. “There are lots of unknowns right now, and real time insights are very valuable right now, as [CMOs] want to be responsive.”
Think about the past few years. Advertisers have fallen in love with - and poured budgets into - ads on retailers’ sites and apps which can connect ad exposures with sales better than anything else.
As a result:
Brands (and their C-Suite friends) want this kind of closed-loop attribution wherever they can find it
Retailers, who are enjoying their boom in ad sales, are looking to extend their capabilities beyond their own properties (so they don’t run out of ways to monetize such data)
Naturally we’ve seen nearly every form of digital media introduce ‘shoppable’ ads and or products
At the same time, we’re seeing retail media data being applied to all sorts of non-shoppable media, from linear TV to OOH - whether media companies think this is a good idea or not
Now throw in a potential economic downturn - or at least an unpredictable consumer landscape - and we may already be seeing an overuse of RM data. While in recent years procurement officers had become convenient boogeymen, this year, it might be those bean counters from finance who want to make everything equally accountable.
“Attribution is the name of the game, no matter what platform,” said Manu Singh, chief data and innovation officer, National CineMedia. “CMOs are getting pressure from CFOs. That is that ecosystem we live in.”
Ok, so potential recession aside, every media dollar is under pressure. And RM data would seem to be a big help - assuming you can track how well each medium drives sales over time. That said, should you really use eCommerce shopping tools to measure ad campaigns that run all over the place, and have different goals to accomplish?
One top traditional media executive recently stated that RM data “is being misapplied, in many places where it has little value, or is unproven.”
It’s understandable that some media executives won’t love being graded on transactions, when it’s just not their company’s strength. That’s not going to stop people from trying to use it as an effort to level up attribution overall.
“I don’t think it’s an overuse,” said Adam Skinner. Managing Director of Unified Retail Media at Epsilon “More like a misuse. When you start over-indexing on specific click metrics, last click, etc., that can have a big downstream impact, and start skewing decisions.”
Skinner and others mentioned how this can result from many brands and agencies are falling back in love with using MMMs (Marketing Mix Modeling), which may not account for retail media’s impact amidst data from brand-building vehicles like TV.
“With MMMs, things can go wrong,” said Miller. For example, linear TV will score high on classic metrics, while RM has different strengths “Reach and effectiveness are two different things.”
That said, if MMMs aren’t the way to go, what about multitouch attribution providers? Well, the danger there, per Skinner, is “when people are looking at traditional media with the same rigor as retail, as if they are a one for one channel. They should not be treated equally.” In other words, you can’t expect a 30-second spot during an NBA game should perform in the same fashion as an ad on Amazon.com - and whatever model you use should reflect that.
So does that mean that brands (and CFOs) should not be pushing to apply RM like standards to out-of-home, TV, even in-store media?
“In reality, there are several issues,” said Singh. For instance, “all of these retail entities are siloed. Trying to apply that data across media plans is complicated, and requires significant coordination, and discussion about what signals matter. Because signals get bastardized, and you can end up arguing about MMM vs. MMA.”
Sounds like a really cool debate!
Naturally, Christine Foster, Senior Vice President, Commercial Strategy & Operations at Kroger’s 84.51 division, sees things a bit differently. “I don’t think we can overuse this signal data,” she said. “Demo targeting in TV is kind of archaic (and shopper data) can be so enlightening, particularly against the ever changing audience behavior.”
Foster sees lots of potential for using RM data for smarter targeting upfront. For example, certain electric car brands might want to target people who spend more on organic products. “You are going to reach people who are most interested [in your products], and it is a shift form the prior way of identifying audiences. But how you use it matters. Not everything that looks like a nail needs a hammer. You need to ensure that you understand each channel’s role. Not everything is going to drive a sales on Kroger.com.”
The question is whether the CFO gets that. In the meantime, brands are testing all sorts of innovative usage of shopping data.
Even cinema ads - which might seem like the ultimate mass marketing branding vehicle, are employing “clean rooms, and household data matchups,” said Singh. “We have a very robust tech stack, and can do outcomes-based guarantees. What cinema has become is that unified shared experience, you are now out and about, walking past all the retail locations. You are in the window.”
Sure, but how do you know how long the purchase window should be for a movie ad, or an out of home ad, or a Netflix show? Miller said these are things brands and partners need to figure out together as part of testing.
“You want to tell a true omnichannel story, and this data can give you a pretty compelling view of consumer behavior. It’s so important to test.”
Right. But who can afford to do that sort of thing if the market becomes as volatile as some think it might. As Epsilon’s Skinner put it, “what brands are trying to do is look for ways to reach more unique customers. So you might see less experimenting, and more, ‘we have a goal, we need to pick that channels that get us to that goal.”
That’s probably good for retail media on its own. But is that good for other media trying to join the accountability party? I can see impatient CFOs, or CFOs pushing for the use of data where it doesn’t make sense.
“Brands are going to go overboard,” said Singh. “But RM data is a bridge, not a magic wand.”