The Drum Awards for Marketing - Entry Deadline

-d -h -min -sec

Agencies Agency Models Data

Dentsu, Publicis, IPG and Omnicom share how they differentiate their data businesses


By Sam Bradley | Senior Reporter

November 15, 2021 | 8 min read

As agency holding companies deepen their investment and dependence on data units, we explore how they plan to stand out to clients as part of The Drum‘s Data Deep Dive.

data center

How can you differentiate a product that‘s inherently binary?

Data businesses are at the heart of every major advertising and marketing group competing today. Consider, for example, the remarks of Omnicom boss John Wren in its third quarter results. He reserved a special mention for the network‘s data enterprise Omni, crediting it with enabling the group to “consistently orchestrate better outcomes for clients and win new business“.

Across the pond, Mark Read of WPP highlighted the contribution of Choreograph, its in-house data offering, to WPP‘s “reshaped offer... combining creativity with technology and data“. And Read‘s former boss Sir Martin Sorrell credited Media.Monks‘ data practice as the engine room of its business.

With the entire marketing industry now selling services that are underpinned, led or driven in some way or another by data, you‘d be forgiven for not knowing how to distinguish between them. Agencies are supposed to be pretty good at marketing themselves, but how do you differentiate a product that‘s inherently binary?

Merkle’s sector focus

For Dentsu, technology partnerships and a focus on the humans at the root of data collection are key. The company can claim to have something of a first-mover advantage in this area. It acquired a majority stake in Merkle back in 2016 (later progressed to 100% ownership in 2020). Dentsu International global chief executive Wendy Clark told The Drum how it hoped to distinguish its offering from its rivals – which, she says, are principally Deloitte and Accenture.

“They‘re worthy competitors. Competition is good, right? We wouldn‘t be anywhere as good as we are if we didn‘t have great competitors,“ she says. Clark points to Merkle‘s partnership with Salesforce and highlights the way in which its martech offering tessellates with Dentsu‘s wider “transformation“ offering around loyalty schemes and e-commerce – an area which she says is “the fastest growing corridor of our business“.

“Merkle‘s data capability is at the heart of our business,“ she explains. “We say that we‘re going to know people better than anyone else. That‘s how we make meaningful progress for our clients, by being the most integrated network with that human identity intelligence and understanding at the core – that comes straight off of Merkle. That‘s how it underlines the entire proposition of Dentsu.“

Acxiom’s integration and additional capabilities

At Axciom, head of international Mike Menzer tells The Drum that it emphasizes its integration with IPG’s other agencies, and its data management capabilities as key differentiators.

Since bringing the company on board, he says: "With any organization investing in new capabilities, one should listen to client demands, understand market trends and stay true to core competencies, Acxiom is no different. IPG has allowed Acxiom to look at expanding capabilities in areas like data & data management, cloud services & analytics, and first-party identity graphs are just a few to mention today.

"With any acquisition, it takes some time to get the engagement model right; we have evolved this with the agencies where we typically work in collaboration and partnership with our sister companies Kinesso and Matterkind – this is and will be increasingly a core differentiator for IPG but also a benefit to our clients. IPG also uses an Open Architecture approach, meaning they field the best team to meet the clients’ needs, and Acxiom, Kinesso, and Matterkind are key to growing and winning those opportunities."

Menzer says the combination is working; he shares that IPG’s Acxiom has seen a 95% contract renewal rate among clients.

Omni’s multitool

One potential means of differentiation can come from the way these offerings were first built. Merkle and Epsilon, for example, were acquisitions, in contrast to Choreograph and Omni, which were constructed and built out in-house.

Jonathan Nelson, chief executive officer of Omnicom Digital, points out that its data offering Omni is not a bolt-on acquisition but a series of tools. “We’re an operating system that sits underneath everything.“

While Omni itself approximately 1,300 engineers working on it, 40,000 members of Omnicom‘s wider staff have been trained to use the operating system in some capacity, with 10,000 of them using it regularly. While Omni itself comprises a hundreds-strong, constantly updated tech stack, Nelson points to its mix of quantitative and qualitative as a competitive edge.

Q, the ’cultural intelligence system’ from Omnicom consultancy Sparks & Honey, is powered by Omni, connecting its vast data gathering and analysis efforts to a human output. “It leans very heavily on machine learning to sift through huge amounts of data coming in... the point of it is really meta-trend analysis. It‘s basically a search tool that will highlight things that are trending... it also has predictive analytics that say where we believe things are going.“

Based on data drawn from disparate sources – not just publisher cookies or Google search habits, but Twitter, the US patent database and the entire archive of The New York Times – Q can spotlight key trends in culture. Meanwhile, Omni helps Omnicom agencies engage in dynamic creative optimization (DCO), the busywork of working out which typeface and which color will persuade the most consumers to buy a certain beer or choose a certain airline.

Says Nelson: “All companies want to sell things. We‘re capitalists, that‘s why we and they exist. But companies also want to build their brand. So you have a quantitative result, which is how many cars we sell, and a qualitative result, which is how did we change perception. Omni is the foundation for all that, and that‘s very differentiated from our peer set.“

Epsilon’s eye on the bottom line

At Publicis Groupe’s Epsilon, chief operating officer Ric Elert is just as keen as Clark to big up the depth and precision of its data collection efforts. “We have an identity capability that is two times better than anything else out there in the marketplace. We‘re able to identify our clients‘ customers not only to activate, but also to personalize and measure.“

It’s the addition to Epsilon’s existing datasets, and the underlying technology used to make sense of it all, that he says has led to the group’s string of recent client wins. “A lot of times it comes down to... [whether] the numbers prove that you can provide extra scale. If you don‘t find more uniques out there in the marketplace, you‘re just going to be over-marketing to a smaller group of people. That doesn‘t drive efficiency.

“Driving more unique customers for our clients is key. For almost every one of them coming in recently, that’s top of mind – how do I maximize the customers I have? How do I acquire more and drive them into our business?“

The third element, Elert says, is “how do I open that up to my manufacturers?“ Its specialism in helping clients directly cash in on the data they already have, rather than just use it as inspiration, is Elert‘s killer edge. While Epsilon‘s primary function is to help clients “better harness their first-party data and augment it... to activate against their customers and drive performance“, he says monetizing first-party data for suppliers is also a selling point.

“We help them monetize that data to their manufacturing partners to drive secondary businesses off clients’ data – it’s working as an agent of the client,“ he says. Epsilon was recently praised by Forrester for its work in this area, especially for retail brands. “It‘s a quickly growing sector... that we‘re a leader in.“

Even as they seek to stand out in the marketplace through focusing on integrated agency work, cultural insights or directly monetizing data, agency rivals are likely to only deepen their investments and dependence on their data units. As Nelson says: “The world‘s not getting less technical or less complex.“

Agencies Agency Models Data

More from Agencies

View all


Industry insights

View all
Add your own content +