The $2.2 billion acquisition that could change how AI agents are built for marketers

Publicis Groupe has announced it will acquire LiveRamp, one of the most widely connected data collaboration platforms in advertising, in an all-cash transaction valued at $2.2 billion. The deal is expected to close by end of year, pending regulatory approval and LiveRamp shareholder sign-off. LiveRamp CEO Scott Howe will continue to lead the business, reporting directly to Publicis CEO Arthur Sadoun. 

On the surface this looks like another large agency acquiring ad tech. But the strategic rationale behind it is worth understanding because it reflects a broader shift in how the industry is thinking about AI, data, and what will actually separate winners from everyone else over the next few years 

What LiveRamp does and why Publicis wants it. 

LiveRamp helps brands connect, manage, and activate data from multiple sources, like CRM platforms, loyalty programs, in-store data, and retail media, in secure clean room environments. The result is what the company calls co-created data: synthesized assets that combine signals from across a brand’s ecosystem into something more useful than any single data source alone. 

For Publicis, that data layer is the missing piece for building more sophisticated AI agents on behalf of clients. The argument is straightforward: agents built on co-created, continuously updating data learn and improve with every signal. Agents built on static or generic data don’t. Publicis is betting that data quality will be the primary differentiator in the agentic AI race and that LiveRamp gives them an edge most competitors can’t easily replicate. 

The agentic AI context behind the deal. 

This acquisition doesn’t exist in isolation. Publicis has been making a series of moves to establish itself as the leading agency network in the agentic AI era. Last month it deepened its relationship with Microsoft around agentic AI while winning Microsoft’s global media account. The LiveRamp deal accelerates that strategy by giving Publicis the data infrastructure to power agents that can span the full customer journey from awareness through conversion and measurement. 

The practical example Publicis outlined in the announcement is illustrative: a retailer could build an AI agent that connects CRM data, loyalty data, in-store data, and retail media inventory, then feeds that back to partners to improve measurement and map the customer journey in ways that weren’t previously possible.  

The neutrality question worth watching. 

Bringing an independent ad-tech platform under a major holding group always raises questions about data access and fairness. Publicis has moved to address that directly. The announcement emphasizes that LiveRamp will remain interoperable, that pricing and commercial practices won’t change, and that no existing or prospective customer will be restricted from accessing its services. 

Whether that holds up in practice as the integration deepens is worth monitoring. LiveRamp’s value to the broader market depends significantly on its neutrality. It’s interoperable with over 25,000 publisher sites and hundreds of technology partners. If that ecosystem starts to feel like it’s tilting toward Publicis clients, the competitive advantages the deal is designed to create could come with reputational costs. 

What this means for B2B marketers. 

The Publicis-LiveRamp deal is part of a larger consolidation trend in marketing infrastructure. As AI becomes central to campaign execution, the agencies and platforms that control the best data will have a structural advantage. For B2B marketers evaluating agency relationships or martech investments, data ownership and interoperability are becoming as important as creative capability or media reach. 

Publicis is also projecting that the deal will raise its net revenue growth outlook from 6-7% to 7-8% for 2027 and 2028, which signals real confidence in the commercial payoff. The Epsilon acquisition in 2019 attracted skepticism at the time and proved out over the long run. This deal is a similar bet on the same thesis: that data infrastructure is worth more inside a scaled network than it is as a standalone business. 

The takeaway. 

The race to build better AI agents in marketing is really a race to build better data foundations. Publicis is making a $2.2 billion argument that co-created, continuously learning data is the moat and that whoever controls it will build agents that outperform everyone else’s. For B2B marketers, the implication is practical. Your data strategy and your AI strategy are not separate conversations anymore. The brands that treat them as one are going to have a real advantage over the ones that don’t.