WPP is putting its agentic AI platform underneath a substantially simplified operating structure, combining technology deployment with corporate reorganisation as the advertising group attempts to reverse declining revenue, remove duplicated agency infrastructure, and change how work moves between creative, media, production, and enterprise teams.

First-half revenue fell 4.4% to £6.37 billion, while revenue less pass-through costs declined 4.7% like for like to £4.75 billion. The second quarter showed some improvement, with the same measure falling 2.8%, although WPP said legacy account losses continued to weigh on performance and its 25 largest clients remained down 6.3% across the half.

Headline operating profit reached £398 million and the associated margin increased slightly to 8.4%, helped by lower severance costs and broader savings despite the revenue decline. WPP expects £100 million of savings during 2026 as part of Elevate28, a restructuring programme intended to produce £500 million of gross annualised savings by 2028.

Financial pressure is now being addressed alongside a structural overhaul that replaces much of the traditional holding-company arrangement with four operating units across four regions. WPP Media, WPP Creative, WPP Production, and the newly launched WPP Enterprise Solutions are being connected through WPP Open, the group’s internal AI and marketing platform.

AI moves into the operating structure

Advertising groups have spent the generative-AI cycle launching tools, establishing technology partnerships, and promising faster content production, often without changing the organisational machinery underneath. WPP’s current programme goes further by making the technology layer part of an attempt to remove internal boundaries, standardise workflows, and present large clients with a more integrated organisation.

WPP Open increasingly handles day-to-day creative, production, and media processes, while its Open Intelligence data layer is intended to connect client and partner information with AI systems used across those workflows. During the second quarter, WPP expanded relationships with Google, Meta, and AWS, including predictive and generative AI integrations that feed into the common platform.

The company also formally launched WPP Enterprise Solutions on 1 July, creating an operating unit intended to capture corporate demand for AI transformation beyond conventional advertising assignments. That brings the group closer to consultancies, systems integrators, and enterprise technology services providers at the same time as those companies are adding more creative and marketing capability of their own.

Generative AI cuts across many of the boundaries on which traditional agency structures were built. A campaign can involve customer data, automated audience selection, generated creative variants, media purchasing, measurement, ecommerce, and production within the same technology-supported process, making historic divisions between agency disciplines harder to maintain when they increasingly share models and data.

WPP is attempting the reorganisation during a commercial downturn rather than a period of easy growth. First-half revenue less pass-through costs at WPP Media fell 5.4% like for like, while WPP Creative declined 4.9%; WPP Production grew by 1.6%, providing a comparatively small area of expansion within the integrated-agency businesses.

The group has reported some improvement in client retention and new business, including work for Estée Lauder, Henkel, Tesco, and Deutsche Bahn, but legacy account losses continue to feed through the numbers. WPP Open therefore has to be judged through client retention, integrated mandates, margins, and delivery costs rather than internal adoption or the number of AI functions available.

Productivity creates another tension. Agency groups historically generated revenue through large workforces performing labour-intensive creative, planning, production, and media tasks, while generative systems can reduce the human time required for some of that work. Productivity gains can improve margins, but they also put pressure on pricing where clients conclude that tasks previously requiring days can now be completed considerably faster.

WPP’s response appears to be a shift towards integrated outcomes rather than selling disconnected agency capacity. Common incentives are being introduced across operating units, WPP Creative has been reorganised into four regional profit-and-loss structures, and Enterprise Solutions is intended to package technology and transformation capabilities alongside the group’s existing marketing work.

The £500 million savings target consequently goes beyond a conventional cost-cutting exercise, although execution will determine how much represents genuinely improved workflows rather than simply a smaller organisation. AI can remove repetitive production and coordination work, but integrated structures can also introduce new governance demands where teams have to reconcile client data, permissions, model choices, brand controls, and accountability across global operations.

Advertising is becoming one of the clearest tests of how generative AI changes professional-services economics because the technology can already perform meaningful portions of content creation, research, adaptation, and workflow automation. Large advertisers still require brand governance, regulatory control, media strategy, data management, and coordination across markets, leaving considerable value in the organisational system around the models.

WPP is betting that its scale becomes more useful when those functions share a common technology layer rather than sitting behind separate agency brands and systems. With revenue still contracting, WPP Open is being asked to support a commercial recovery, organisational simplification, and a different delivery model simultaneously.

The next evidence will come through client economics. If shared technology and fewer internal boundaries improve retention, integrated wins, and delivery costs, WPP will have a credible example of AI changing a major professional-services operating model; continued revenue decline would make widespread AI adoption a much less persuasive measure of progress.