The Numbers That Explain the Advertising Agency Industry Restructuring of 2025
WPP lost 9,400 staff in a single year. Omnicom absorbed IPG and immediately announced 4,000 redundancies. Dentsu posted a record net loss of ¥327.6 billion ($2.18 billion USD). These are not cyclical corrections. They are the visible surface of a structural shift that has been building for years, and the advertising agency industry restructuring of 2025 marks the point where that pressure became impossible to manage quietly.
For businesses that rely on agency partnerships, or are evaluating them, the change matters beyond the headline numbers. The firms being restructured are the same ones that dominate global pitch rosters. How they are changing, and why, directly affects what you should expect from an agency relationship in 2026 and beyond.
What Is Actually Happening at the Holding Companies
The three major holding group stories of 2025 share a common thread: the scale that once justified their existence is now functioning as overhead.
WPP entered 2025 having already posted a 1% like-for-like revenue decline in 2024. It finished the year down 5.4% on revenue less pass-through costs, with 98,655 people on the books after starting the year at 108,044. That is roughly 9,400 jobs gone in twelve months. In February 2026, incoming CEO Cindy Rose (formerly of Microsoft) announced the Elevate28 strategy, collapsing the group into four divisions: WPP Media, WPP Creative, WPP Production, and WPP Enterprise Solutions. The cost savings target is £500 million per year, with programme cash costs of around £400 million phased over two years.
Rose’s framing of the strategy was notable for its candour. “We don’t want to be a holding company any more,” she told investors and press on 26 February 2026. That is a significant statement from the CEO of the world’s largest agency group. It reflects a recognition that the holding company model, in which loosely connected agencies share a parent balance sheet but operate independently, no longer serves clients or the business.
WPP’s client losses made the structural fragility concrete. L’Oréal moved its media to Publicis in 2024. Coca-Cola’s $700 million US media account followed. Mars shifted its $1.7 billion media budget to Publicis. PayPal’s global media account went the same way. Starbucks moved to Stagwell’s Anomaly in January 2025. The pattern is not random account churn. It is a directional signal about where marketers believe capability sits.
Omnicom and IPG merged at end-2024 after FTC and EU clearance. The combined entity entered 2026 with 120,000 staff and a restructuring savings target that the group doubled from $750 million to $1.5 billion, comprising $1 billion in labour costs, $240 million in real estate, and $260 million in IT and procurement. The merger also retired three major creative brands that had defined the industry for decades: DDB and MullenLowe folded into TBWA and McCann respectively, FCB was absorbed by BBDO. The 4,000 redundancies announced on 1 December 2025 came on top of approximately 8,200 roles shed between the two groups across the year.
Dentsu’s situation is more acute. The group’s record net loss of ¥327.6 billion was driven primarily by a ¥310.1 billion goodwill impairment on its international operations, representing a formal accounting acknowledgement that its global expansion, which involved decades of acquisitions, is worth substantially less than it paid. International revenue fell across every region: Americas down 3%, EMEA down 1.8%, APAC ex-Japan down 6.8%. The company attempted to sell its international operations and found no buyers. Its share price fell approximately 11% on that news. Dividends have been suspended for both FY2025 and FY2026. The group is cutting 3,400 roles internationally, around 8% of its international headcount.
Why This Is Structural, Not Cyclical
Agency revenues are sensitive to economic cycles, so it would be easy to attribute all of the above to a tough advertising market. But the numbers do not support that reading.
Global ad spending grew in 2025. The holding companies did not. The divergence between aggregate market growth and holding company performance points to something more fundamental: share loss. Advertiser Perceptions data shows the holding companies’ combined share of total US ad spending fell from more than 44% in 2019 to 29.6% in a recent quarter. That is not a recession effect. It is structural erosion.
Three forces are compressing the model simultaneously.
The first is AI. The tasks that once justified large teams at agencies, particularly in media planning and buying, are increasingly automated. Audience clustering, lookalike hypothesis generation, real-time budget optimisation, report drafting: these were billable hours. AI tools now execute many of them in real time. Forrester data shows agencies averaged 8% headcount reduction in 2025, with a further 15% reduction forecast for 2026, driven by AI and structural pressures combined. Ogilvy cut approximately 700 roles in June 2025, about 5% of its 14,000 global staff, citing AI integration alongside broader restructuring. Horizon Media cut 50 roles from its global team in March 2026, with CEO Bill Koenigsberg specifically citing AI and calling it a “skills optimization effort.”
The second force is in-house migration. According to the ANA’s 2024 In-House Agency Fact Book, 82% of ANA members now have in-house agencies. In 2008, that figure was 42%. The trajectory is not ambiguous. Of the companies that have built in-house capability, 69% reported cost savings of 6% or more, and 48% reported savings exceeding 20%. Gartner’s 2025 CMO Spend Survey found 39% of CMOs planning to reduce agency allocations, and 22% saying that generative AI had already enabled them to reduce reliance on external agencies.
The third force is scale itself. The holding company model assumed that aggregating thousands of people, dozens of brands, and global infrastructure created value. For clients who wanted global reach and category depth, it did. But AI-assisted production, distributed creative platforms, and senior independent practitioners have eroded those advantages significantly. At the same time, the complexity of managing a global holding group, including the coordination costs, management layers, and internal politics, has become visible to clients. Agencies billing at holding company rates while routing briefs through multiple layers of junior staff are a harder sell than they were a decade ago.
The Publicis Divergence and What It Signals
Not every major agency group is in the same position. A separate article covers the Publicis story in depth, but the contrast is relevant here.
Publicis grew organic revenue 5.6% in 2025, approximately 700 basis points ahead of the peer average, which the company described as expected to be negative overall. It added approximately 5,800 staff while its peers shed thousands. It won 56% of all global new business billings tracked in 2025, double Omnicom’s 656 wins. Its market capitalisation exceeded €26 billion at end-2024, having more than doubled over five years.
The contrast reflects a decision made in 2017, when Publicis invested heavily in an internal AI platform called Marcel before AI was fashionable, and expanded it through the CoreAI initiative announced in January 2024. The €300 million, three-year CoreAI investment brought together 2.3 billion consumer profiles, decades of transformation data from Publicis Sapient, and all existing Marcel assets. The industry mocked the Marcel bet when it was made. Seven years later, the gap it created is showing up in market share data.
The point for businesses is not that Publicis has solved everything. It is that the AI divide within the agency industry is already creating divergent performance, and that divide will widen as AI capability compounds. A group growing 5.6% and one shrinking 5.4% are not operating in the same market, despite competing on the same pitch rosters.
What This Means If You Are Choosing an Agency Partner
The structural shifts above create practical implications for any business evaluating agency relationships.
The first implication is that size is no longer a reliable proxy for capability. For years, a large agency holding company brand on a credentials deck signalled access to global talent, best-in-class tools, and category expertise. Those signals still carry some weight, but the firms most invested in AI-driven delivery are outperforming the largest ones. The better question for a brief is: where does the actual work get done, and who is doing it?
The second implication involves continuity risk. Agencies undergoing restructuring, whether shedding brands, merging teams, or cutting 8-10% of their staff annually, are not stable operating environments for client accounts. Senior talent at restructuring firms is actively being recruited by independents and in-house teams. The person who pitched your account may not be running it six months later.
The third is about what AI actually changes for clients. Advertiser Perceptions found 52% of US agencies and marketers agreed AI will devalue the role agencies play in media planning and buying. That devaluation benefits clients directly: tasks that previously justified large retainers are becoming automated. Businesses that understand this have leverage in how they structure and price agency relationships. It also means the value an agency should be delivering has shifted toward strategy, judgment, and implementation quality rather than production volume.
For mid-market businesses without the budget or appetite to build the in-house capability that large corporations are investing in, the calculus is changing. The gap between a senior-led independent agency with modern tooling and a bloated holding company network, managed through layers and billed at premium rates, has narrowed considerably. In some categories it has reversed.
Evaluating Agency Relationships in a Restructuring Market
The advertising agency industry restructuring of 2025 is not a story about which large firms are having a bad year. It is a signal about what the agency model is becoming, and what it is no longer. Scale is being replaced by specialisation. Broad services networks built on headcount are being reshaped by AI compression. Clients who built in-house capability over the past decade look well-positioned. Those still renewing large holding company retainers on the assumption that structural advantages justify the premium deserve to ask harder questions.
The businesses best placed to navigate this shift are those that align their agency partnerships to actual capability rather than brand legacy, and that understand how AI is reshaping what agencies should be charging for.
Avatar Studios works with businesses across AI and automation, digital products, growth optimisation, and strategy. If you are reassessing your agency model, or looking for a direct, senior-led alternative, the conversation starts at avatarstudios.com.au/services.