Publicis Isn't Winning Pitches Anymore. It's Skipping Them.

PepsiCo's new "One PepsiCo" media model, unifying strategy, planning, activation, identity and technology across Pepsi, Gatorade, Lay's and everything else the company sells, will run through Publicis Groupe in more than 200 markets. It displaces Omnicom's OMD network, which had held pieces of the account, including the US and UK, for more than two decades. And according to Adweek's reporting, there was no pitch. PepsiCo appointed Publicis "following a media capabilities review."

Sit with that phrase for a second, because it's doing more work than the headline. A capabilities review isn't a competition, it's closer to due diligence on a decision the client has effectively already made. That distinction is the honest starting point for answering why Publicis appears to be winning almost everything right now, and it isn't the whole answer, because Publicis has also been winning plenty of accounts through real, contested pitches. Kering, the parent of Gucci and Balenciaga, ran an actual competitive review for its global media account this July, managed by the independent pitch consultancy ID Comms, and Publicis won that one outright against real competition. Add L'Oréal, McDonald's, Samsung, Procter & Gamble, and, from the US market specifically, Kenvue, Mars, PayPal and TJX, and CEO Arthur Sadoun told investors in July that six major new clients in the first half of 2026 alone added €300 million in net new revenue, enough to add roughly two percentage points to the group's full-year growth. That's not a fluke streak. It's sustained enough that a genuine "why is Publicis winning everything" thread has been running on advertising subreddits and trade forums for months, people inside the industry are asking this question too, not just outside observers.

Why the others aren't matching it

The uncomfortable answer for Publicis's rivals is that this isn't really about who pitches better in the room. WPP has spent the past two years restructuring into four divisions, changing CEOs, and publicly trying to convince the market it's not a holding company anymore, while its share price roughly halved over the past year and its own outcome-based pricing pivot, meant to be the centerpiece of its turnaround, has so far landed exactly one client. That's not a company positioned to win a global capabilities review right now; it's a company still explaining its new org chart internally. Omnicom, meanwhile, has spent 2025 and 2026 absorbing its acquisition of IPG, a genuinely enormous integration task that includes dissolving IPG's Kinesso data unit and Omnicom's own Annalect into a single combined structure. Both of those companies have real, understandable reasons for their current internal focus. Neither of those reasons is "we didn't have good enough people in the pitch room." Publicis, notably, wasn't doing either of those things over the same period. It was doing something else: compounding.

Since 2017, Publicis has been building Marcel, its internal AI platform, years before "AI-powered" became a mandatory line in every holding company's earnings deck. It's put roughly €12 billion into data and technology infrastructure over the past decade, and this year added LiveRamp, a $2.2 billion identity-data acquisition, to a list that already includes Epsilon. When PepsiCo asked for a single, AI-and-data-driven operating model spanning 200 markets and an entire brand portfolio, Publicis wasn't proposing to build that. It already had the infrastructure sitting there, tested across other clients, ready to point at. That's the actual mechanism behind "no pitch." You don't run a competitive review when only one bidder can demonstrate the thing you're asking for already exists and already works.

The client-concentration question the article itself already answered

Here's where the PepsiCo story gets more interesting than a simple win column, and where the client's own instinct about attention and bandwidth turns out to be exactly right, just not in the way it might expect. The moment Publicis took PepsiCo, it withdrew from the ongoing pitch for the rest of Coca-Cola's global media business, an account MediaSense has been running a review for, reportedly worth around $4 billion. Publicis already handles Coke's US and Canada media, won earlier from WPP. It chose not to try to hold both Pepsi and Coke at the global level simultaneously.

That's not a regulatory requirement forcing Publicis's hand. Agencies inside holding companies serve competing categories all the time through internal conflict walls and separate teams. This was a choice, and it's the most honest data point available on the question of whether a holding company stacked with major clients can genuinely give a new account enough attention. Publicis itself apparently doesn't believe it can run PepsiCo's full global media model and compete seriously for Coca-Cola's global business at the same time, at least not without compromising one or both relationships enough that it wasn't worth trying. If the company best positioned to prove client-stacking is manageable won't even attempt it for two of the largest, most direct rivals in the world, that's a stronger signal than any reassurance a new business team could offer a prospective client. The honest question every brand should be asking Publicis right now isn't "do you have room for us," it's "which of your current relationships would you drop before you'd shortchange mine, and how do I know."

What this actually means for everyone else

For the other holding companies, the uncomfortable lesson from PepsiCo isn't "pitch harder." It's that restructuring an org chart, which is what WPP has spent two years doing and what Omnicom is currently doing by necessity, doesn't produce the same outcome as owning infrastructure a client can't quickly replicate elsewhere. A four-division reorganization is something a competitor can read about and copy the shape of within a quarter. A decade of accumulated, tested data and identity infrastructure is not. Catching up here isn't a messaging problem to fix before the next pitch season, it's a multi-year investment decision that has to start now to matter in three years, and neither WPP nor Omnicom is in a financial or organizational position to make that bet at Publicis's pace while they're mid-transformation of their own.

For independent agencies, the opening is narrower but real, and it's the same one running through Publicis's own client-concentration dilemma. An independent will never win a 200-market, full-portfolio global media consolidation; it doesn't have the infrastructure and shouldn't pretend to be pitching for that job. But it can credibly make the opposite promise a stacked holding company structurally can't: undivided attention, no sister-brand conflict to manage, no need to ever choose between two of your competitors because it was never trying to serve both in the first place. Publicis just demonstrated, by its own actions, that this attention question is real enough to walk away from $4 billion over. That's not a weakness to exploit cynically. It's a genuine, evidenced trade-off clients are entitled to weigh, and the agencies that win the accounts Publicis structurally can't take will be the ones that make that trade-off explicit instead of pretending scale and undivided focus can both be true at once.