The Moat No Merger Can Buy

You don't need or want any more commentary this week about how WPP hired McKinsey to advise on its strategic review, or that Havas is reportedly exploring some kind of arrangement with WPP, or that Omnicom and IPG are weeks away from closing their $13.25 billion merger.

All of these stories share a common thread though: they're attempts to solve a structural problem that can't be solved through acquisition, merger, or outside consultants.

The problem is simpler and more fundamental than most realise. WPP, Omnicom, IPG, Havas, and Dentsu are HOLDING companies. While Publicis is a company. And Publicis is winning.

That distinction of HOLDING company versus company represents the widest competitive moat in the agency business. And it's the one moat that cannot be bought, merged, or restructured into existence overnight.

I spent thirteen years at Publicis, six on the Management Committee where alongside Arthur Sadoun and the leadership team we executed what might be the most underappreciated transformation in modern agency history (it certainly was at the time, it was ridiculed even, from outside)… From 2017 to 2020, Publicis systematically dismantled the holding company structure and rebuilt itself as an integrated company.

This wasn't rebranding. This wasn't a press release announcing "One [Company Name]" while leaving all the silos intact. This was grinding, painful cultural work that required blowing up the traditional holding company model entirely.

Here's what actually changed: Publicis eliminated agency brand P&Ls (why? To solve a murder, start by following the money). The primary P&Ls became geography and client accounts. Leo, Saatchi & Saatchi, BBH, Starcom, Zenith, these brands still exist, but they're signatures within integrated client teams, not independent business units with separate leadership, incentive structures, and tribal identities. They exist for clients when they create value, otherwise not.

The 'Wolf Pack' is a common leadership metaphor trope. It usually goes like this: Alpha wolves lead through dominance. Subordinate wolves compete for status. The pack succeeds through rigid hierarchy and competitive pressure.

Except that's not how wild wolf packs actually work. That model comes from studying captive wolves where unrelated individuals are forced together in zoos and laboratories. When researchers put stranger wolves together artificially, they fought for dominance, established rigid hierarchies, and competed viciously for resources.

Researcher David Mech, who popularised the "alpha wolf" concept in the 1970s, spent 13 summers observing wild wolves in their natural habitat. His conclusion? He was wrong. In nature, wolf packs aren't hierarchies. They're families.

A wild pack is a breeding pair and their offspring across multiple generations. The "alphas" are just parents. The "subordinates" are children and teenagers. There's no dominance fighting because parents don't need to fight their own children for status. Knowledge transfers naturally across generations where older wolves teaching younger ones hunting techniques, territory management, and pack cooperation.

The pack cares for injured members. They slow travel pace for elderly wolves. They feed sick pack mates. Research shows wolves form lifelong bonds and show evidence of grief when pack members die. As the Living with Wolves organisation states: "Everything in a wolf's nature tells it to belong to something greater than itself: a pack."

Here's where it devastates the holding company model:

Traditional holding companies are using the captive wolf model where unrelated agencies forced together under one corporate banner. WPP didn't grow VML and Ogilvy together as siblings. They acquired them separately and told them to cooperate. Omnicom didn't raise BBDO, DDB, and TBWA as a family. They're distinct agencies with different cultures, competing for resources while sharing a parent company.

When you force unrelated wolves together in captivity, they compete for dominance. When you force unrelated agencies together, they compete for P&L credit, protect their margins, and prioritise agency brand success over holding company objectives.

Each agency maintains separate systems, tools, and workflows, not because of technical limitations, but because integration threatens autonomy. Agency presidents are compensated on their brand's performance, creating zero incentive to sacrifice margin for integrated client solutions. Moving talent between sibling agencies is treated like hostile poaching rather than strategic deployment.

This is the structure that creates permanent friction.

The Publicis transformation took three years of grinding cultural work because you can't fake family dynamics. Every agency leader had to surrender autonomy. Every creative director had to collaborate with former competitors. Every employee had to shift identity from "I'm a Leo Burnett person" to "I'm a Publicis person who works in creative."

This sounds simple to state. I cannot overstate how hard it was to achieve. Agency leaders who built careers running independent businesses had to surrender P&L autonomy. Creative chiefs had to collaborate with competitors from sibling agencies. Compensation structures had to reward client outcomes over agency brand profitability.

This is the transformation that took three years of grinding cultural work. And this is why Publicis operates fundamentally differently than every competitor, in a category of one.

WPP's decision to hire McKinsey for strategic advice reveals the problem. McKinsey can recommend structural changes, portfolio optimisation, and efficiency initiatives. But McKinsey can't change culture. McKinsey can't eliminate the deep tribal loyalties that make VML and Ogilvy function as separate companies despite sharing the WPP parent.

(Worse, in all of the blue-hued charts that are being presented to WPP leadership even now, they probably won't even note this is even a problem. The widest strategic moat is so misunderstood and undervalued.)

WPP hiring McKinsey is like hiring a biologist to make captive wolves cooperate. The biologist can recommend territory divisions and feeding schedules. But they can't create family bonds. Those require years of hunting together, caring for each other's injured, and raising young together.

The Omnicom-IPG merger also demonstrates this reality. With 20,000+ job eliminations expected to deliver $750 million in synergies, the companies are pursuing the traditional holding company playbook: consolidate back-office functions, eliminate duplicate roles, and maintain agency brand identities.

But they're not eliminating the fundamental friction. Each silo will protect its P&L. Each will compete for talent and resources. Each will prioritise agency brand outcomes over client outcomes when those interests conflict.

Omnicom and IPG are 3-5 years behind Publicis on operational integration, not because they're poorly managed, but because they haven't made the fundamental choice to stop being a holding company and become a company. That choice requires abandoning the agency brand P&L structure that has defined holding companies for decades.

You can acquire wolves. You can't acquire pack dynamics.

It's not hard to break down the Publicis results. They win more clients, they keep more clients they have, and they make for from each client. In an increasingly complex marketplace, clients consolidate with partners who deliver integrated solutions without organisational friction. When Coca-Cola moved their North American media business from WPP to Publicis, they were buying elimination of the friction they experienced trying to coordinate between WPP agencies.

For clients, integrated company structure means:

Seamless Access: Epsilon's data, Sapient's technology, Starcom's media, Leo Burnett's creative, all within unified teams, not separate pitches.

Aligned Incentives: Account teams compensated on client outcomes, not agency brand P&L protection.

Talent Fluidity: Best resources deployed without internal politics or transfer negotiations.

Technology Integration: Unified platforms, not agency-specific tools requiring complex integrations.

Knowledge Transfer: Multi-generational learning like wild wolf packs with Epsilon data scientists teaching Sapient technologists, creative teams learning media strategy, not captive wolves protecting territory.

Here's why this moat can't be bought:

Time: Publicis's transformation took three years of sustained effort under consistent leadership. Arthur Sadoun has executed the same strategy since 2017 with no pivots, no agency brand reshuffling, no "back to basics" reversals. Competitors lack that continuity.

Leadership Clarity: Sadoun subordinated agency brands to the integrated model and maintained that position through inevitable resistance. That requires conviction that survives through quarters when individual agency brand performance suffers for the greater integrated outcome.

Cultural Surgery: You can't half-transform. Maintaining agency brand P&Ls while announcing "One Company" integration is worse than doing nothing, it creates organisational confusion and cynicism. Full transformation means eliminating structures that built many leaders' entire careers.

Competitive Pressure: Every day spent on internal integration is a day competitors are winning accounts. WPP, Omnicom, IPG, Havas, and Dentsu are simultaneously trying to maintain business development momentum while executing fundamental structural change. That tension makes transformation exponentially harder.

The Family Bond: Wild wolf packs care for injured members. They slow down for elderly wolves. They feed sick pack mates. You can't mandate that behaviour in captive wolves, it emerges naturally in families built over generations.

The bifurcation is accelerating. On one end: Publicis and Accenture with genuinely integrated operations, the natural wolf pack model. On the other: specialist agencies with focused expertise amplified by AI, small family units with deep bonds.

In the middle: traditional holding companies carrying organisational friction they can't eliminate through mergers, acquisitions, or consultant recommendations or the captive wolf model based on debunked science.

WPP hiring McKinsey, Havas exploring WPP deals, Omnicom-IPG merging… these are all attempts to solve the wrong problem. The problem isn't strategic direction, market positioning, or efficiency optimisation. The problem is organisational structure.

Holding companies can't become integrated companies without abandoning the agency brand P&L model that defines them. That requires the kind of cultural transformation Publicis executed from 2017-2020. It's not impossible. It's just exceptionally rare for large organisations to execute successfully.

This is why Publicis operates in a category of one. Not because of data, not because of media scale, not because of technology capabilities, though they have all three. But because they're the only holding company that stopped being a collection of captive wolves competing for dominance and became a family that hunts together.

Publicis built a natural family. Everyone else is running a zoo.

And in an increasingly complex world, less friction wins.

Brandflow is written by Justin Billingsley, who has spent his career on all three sides of the industry's table: senior client, global agency leader, technology founder. First published 18 November 2025 in the Brandflow newsletter on LinkedIn.