From media margins to machine learning: WPP Media’s big bet

“We now have more data than any of our competition reaching five billion consumers, connecting intelligence of all forms, including IDs, across a vast federated ecosystem,” said Lesser on the call. 

It’s a line that’s been making the rounds with clients in the U.S. and the U.K. for months, ad execs told Digiday. 

“They were talking a lot about de-duplicated reach in that meeting,” said a senior marketer, who anonymously shared details of their discussion with WPP.

Behind the jargon — talk of “de-duplicated reach” and “federated ecosystems” — is a larger shift. WPP, like its peers, is trying to move upstream — away from time sheets and media margins toward something closer to software.

“The plan is to move away from selling people’s time, services and their own inventory and move to selling platform-based solutions,” said a former WPP Media exec.

The idea isn’t new, but AI gives it more weight — allowing agencies to package tools CMOs can buy rather than hours they can bill. 

For WPP Media, that tool is Open Intelligence, the group’s data solution and AI model. They power Open Media Studio, WPP’s end-to-end media planning, delivery and measurement platform within the WPP Open services suite.

Unlike Publicis, which owns Epsilon, or Omnicom, which owns Acxiom, WPP doesn’t own a large first-party data set. Instead, Lesser’s bet is that ownership is overrated — that access, scale and the ability to integrate partners is enough. 

Why? Open Intelligence combines the same volume of traditional identity data as competitors with hundreds of additional partners including media, commerce and data companies such as Google, Amazon and TikTok. These datasets — often behind walled gardens — are connected via WPP Media’s federated learning technology allowing the group to analyze and model performance without centralizing or owning the data.

The promise, according to one marketer and three consultants who have heard the pitch: more accurate predictions before a dollar is spent that lead to better outcomes once they are. 

“We have seen success with this already on some of our long-standing clients in terms of outperforming traditional identity-based approaches to performance,” said Lesser on the earnings call. 

That pitch helped win Amazon’s media business in Europe — not because WPP owned a mountain of data but because it convinced Amazon it didn’t need one, said the former WPP exec. Whether other advertisers, especially those less technically fluent, will buy in is less certain. Even Lesser acknowledged as much.

“In terms of Mars, we were in the process of deploying Open Media Studio for [them],” he told analysts on the call. 

It didn’t stick. Mars chose Publicis instead. 

What won out, according to two people familiar with the decision, was Publicis’ ability to show how it could tie creative decisions directly to performance — through a tech stack Epsilon helped build and Publicis actually owns. That kind of vertical integration isn’t something WPP can yet claim. 

And that’s the rub for WPP: it’s asking CMOs to buy into a big idea before the proof is there. That can work if the vision is sharp enough — and clearly it was for Electronic Arts, Hisense and Hero Motocorp, which have all hired or re-hired the group this year. But not every advertiser is sold. They’ve heard similar promises before only to get optimization for clicks and conversations rather than business outcomes. 

What they want is integrated creative and medium backed by credible modelling — not another round of recycled attribution math.

“I came away from a recent meeting with [WPP Media leadership] with a lot of questions and told my contact that I needed to go away and think about it,” said the senior marketer who met with the group earlier this year. 

At the center of that hesitation: de-duplicated reach. Still one of advertising’s biggest unsolved problems. The promise from WPP was that Open Intelligence could deliver a sharper, unified view of how many real people actually saw an ad across platforms. 

“That’s quite the claim and that’s before you consider the fact that they don’t own a lot of data,” said the marketer. 

A lot of that depends on InfoSum — now part of WPP’s stack — and its ability to match and model data across walled gardens and other online spaces alike. During the meeting, and on the call, it was positioned as a foundational piece. 

That centrality is now shaping WPP’s wider strategy. Digiday understands WPP is (partially at least) pulling up the drawbridge when it comes to its third-party tech and data relationships in favor of its internal toolset and select partners that meet specific criteria. According to sources, this has been characterized by WPP rebuffing partnership pitches, with executives there understood to be shifting all media planning and buying to its own suite of products, such as Open Media Studio. 

Still “pulling up the drawbridge” doesn’t mean going it alone. Open Intelligence recently added 22 new partners including FreeWheel, Lumen Research, DICK’s Sporting Goods and Criteo. 

The long road to transformation

In fairness, it’s been just four months since Lesser rolled out his plan, so there’s still time for it to take hold. There are even signs of progress: in some marketers and categories, WPP has been more aggressive on pricing and more attentive to clients it wants to keep, according to one ad exec familiar with those efforts. 

“Based on what I’m hearing from my clients, WPP Media is doing a decent job,” said the ad exec. “OK, they’re not necessarily competitive but they’re making it clear they’re not going to be outmuscled by rivals on pricing.” 

Granted, that’s not necessarily central to the transformation Lesser envisions for WPP Media, but it matters. Especially when the group’s data and AI pitch hasn’t yet been tested in sustained, head-to-head battles with rival holdcos. And given the business’s precarious spot, a win’s a win. 

Revenues fell 4.3% in the first half of the year, down to £6.63 billion from £6.92 billion. The second quarter alone was down 5.8%. 

For a company that still leans heavily on media, those numbers point to something more structural. WPP Media is overdue for a course correction. 

The recent losses of the Mars and Coca-Cola North America accounts alone have pulled an estimated $2.5 billion in billings from the business, according to COMvergence — though WPP has also picked up roughly $1.4 billion so far this year in new and retained clients. 

“Any time you transform a business as dramatically as we have, you’re going to experience some disruption,” said Lesser. “We did see that in the first half but we’re substantially through that transformation.”

Maybe. But the first half left scars. And they’re still fresh. 

Just weeks after stepping into the CEO role, Lesser made waves when he told Adweek “no human would touch a media plan within five years”. The comment — meant to underscore WPP’s AI ambitions — rattled internal teams. The broader point was about creating new roles, not eliminating them but the message got lost. And it didn’t help that months later, Lesser oversaw the merger of performance marketing unit Nexus with the investment team into a new group: Media Management & Delivery. 

Internally, that move raised eyebrows, according to three separate sources. Some saw it as an overdue elevation of Nexus. Others were more skeptical — reading it as groundwork for training AI systems to eventually take over parts of the media buying process. The changes affected nearly half of the business in North America, and triggered a wave of senior departures. 

“It’s fair to say that our main contacts there aren’t comfortable with these changes because of that word “automation” — it’s so prevalent in everything there at the moment,” said an ad exec, who works closely with WPP Media. 

The same upheaval unsettling WPP Media’s North American teams is central to its global reinvention. It’s the market with the richest prizes — and some of its costliest losses — making it both the jewel and the fault line in WPP’s transformation.