UK-based agency holding group WPP today announced it has acquired data clean room provider InfoSum for an undisclosed fee. InfoSum will become part of WPP’s media arm GroupM, which is headed up by former InfoSum CEO Brian Lesser. As part of GroupM, WPP says InfoSum will “power the creation of a new generation of AI-enhanced marketing solutions” for its clients.
InfoSum, founded in 2015, is a major player in the data clean room space, with a particularly strong client base in the UK and Europe. Its technology enables buyers, sellers, and other relevant parties to bring together their separate datasets for things like campaign measurement and targeting, without the need to actually share that data with each other. A number of major sell-side companies including Channel 4, ITV, Netflix, News Corp, and Samsung Ads use InfoSum’s technology, as well as data partners including Experian, TransUnion, and Circana.
WPP says that following the acquisition, InfoSum will be integrated into WPP Open, its company-wide ‘operating system’, enabling clients to combine their own first-party data with data signals available through the InfoSum network, as well as WPP’s own data assets and GroupM’s media intelligence.
InfoSum’s infrastructure will remain interoperable with existing platforms and partners, according to WPP, and CEO Lauren Wetzel will continue to head up the company, while also taking on the role of chief solutions officer for GroupM.
An identity solution for a post-identity world
The acquisition comes at a time when agency groups are spending heavily to grow their in-house data capabilities. Just a few weeks ago, French group Publicis acquired data and identity business Lotame, following its 2019 acquisition of Epsilon. Meanwhile Omnicom’s impending merger with Interpublic Group is driven partly by the prospect of combining Interpublic Group’s data business Acxiom with Omnicom’s own digital commerce business Flywheel.
But the holding groups aren’t all approaching data the same way. GroupM CEO Lesser says his company is looking to build AI-based solutions, which won’t be reliant on traditional ID solutions.
“Traditional ID solutions like those grounded in email addresses only learn from overlapping data points, relying on outdated lookalike models that limit insights,” he said while outlining GroupM’s strategy earlier this year. “Using technologies like federated learning, we can create shared knowledge and predictions across all of our partners without sharing raw data, and activate via a simple connector.”
At a glance, this strategy might look like it sits at odds with today’s acquisition, since InfoSum can be used to identify overlapping data points in the way described above. But as Lesser says, GroupM is betting on ‘federated learning’, a machine learning technique where different entities work together to train a model without sharing their source data with each other. That’s where InfoSum’s technology, which allows data collaboration without data sharing, comes into play.
So while InfoSum will still be used for one-to-one data collaborations for clients, WPP sees further value in using InfoSum’s technology to build AI-based solutions which will ultimately be informed by a massive number of data sets from both WPP and its clients.
“Directly integrating InfoSum’s global data network and technology infrastructure will allow our clients to create even more value from their first-party data and enable us to train client AI models against the most data, from the most places, at unprecedented scale and speed,” said Lesser in a statement. “Our approach recognises the importance of identity data to today’s marketing strategies while allowing us to take advantage of the limitless opportunities for growth we can create by moving beyond them. As more and more clients leverage our AI-first solutions, every client model, every audience, and every campaign will benefit from network effects that will exponentially increase their intelligence and competitive advantage.”