The Open Marriage

Tab four of the RFI. The conflicts declaration. Somewhere past midnight, the cursor sitting on a sheet that asked us to list every client we served in the prospect's category, by market, by brand, by agency.

I ran global new business for Publicis Groupe for years, and that sheet was the one I always paused over. Not because the answer was hard. We always answered it completely, because a conflict discovered later is fatal and a conflict declared early is merely a conversation. The pause was about the question, and whether the intermediary or the prospect was evolved enough to be asking the right version of the question.

Three things ran through my head every time. Why are they asking, and through which lens will they read the answer? How do we disclose where we work with their competitors without breaching the confidentiality of those competitors, who are also our clients, in which case we've failed at the very thing the question is testing? And how do we declare everything without making a solvable situation look unsolvable before the process that solves it has even begun?

Because the honest answer to "where do you work with our competitors?" was a method and not a list. And whether we got to explain the method depended entirely on whether the person reading tab four believed in a word.

In or out. On or off. Faithful or not.

The industry has spent a fortnight arguing over it, because Publicis took PepsiCo's global media without a pitch, stepped out of Coca-Cola's global review the next day, and Coca-Cola then put the North American media account Publicis already held up for review. I wrote about the mechanics last week. The review of one, the rival's options, the decision that wasn't a conflict. I am not going to write that again. The cola story is now over-reported, and it has done its job. It shone a light on a word.

But nobody in the coverage I read has sat where the word gets negotiated: exclusivity was the word clients used. It was never the thing agencies did. Clients asked for a marriage. What they got, every time, was an open one, and everyone in the room knew it.

Because one of the core functions of that job was to manage conflict, elegantly and honestly, and the word made that harder, not easier. The real discipline had a real name. Conflict management. And it ran on a ladder.

Every scale client walked in at the top rung. Not one of them offered to pay for it.

The top rung was total exclusivity: the whole Groupe, or at least a whole "solution" of it, walled off from the client's category or vertical, everywhere. The agency's best case was the bottom rung: no restriction at all. The negotiation lived between them, and it always settled on a rung, never at either end.

Rung two, counting from the bottom: named individuals. The people who would actually think about the client's business, contractually kept off named competitors or off the vertical. Rung three: a named internal team, a bespoke unit with its own name, its own floor, its own permissions, whose members could not work on other clients in the category. Rung four: a whole agency brand, a Leo Burnett or a Starcom, restricted, with the geographies spelled out market by market so that "global" meant a list, not a word. Rung five, rarely and reluctantly, the top.

Notice what every rung actually protects. People. The strategist who understands your business, the planner who knows your calendar, the creative who has heard what you're launching. The ladder is a set of walls around human beings, built at a time when the agency's value was human beings, fractionally owned. One thing I always used to want as a client (and a story I would also tell prospects and clients from the other side of the table) is that the idea you want from your creative leader is the one they have in the shower, or on a walk, or simply staring out of a window. Not the one they have at a desk. Managing conflict was about ensuring that you were the client they thought about first.

Now notice what the client paid for it. Clients believed the size of their business was the premium. And for a long time that was right: a scale account paid for a floor, a team, a P&L, and the wall came with the building. And there was a time when these clients literally built out agency networks. Saatchi & Saatchi flags on the map were planted by P&G from the 1980s onward; that client "turned the lights on" each day across the network, and exclusivity was carefully protected.

But the wall was never free to the agency. What it cost was the addressable market, every future client in that category, gone for the length of the contract. So we priced it. Just never in money. To my recollection, what we asked for in return for climbing the ladder was time. Longer notice periods. Exit penalties. A defined process for leaving, with steps, and dates, and consequences. Five weeks ago, in The Skin Game, I wrote that the industry already runs on outcomes-based compensation: "The outcome being priced is trust. The payment is tenure. The penalty clause is the review." Exclusivity was paid for in the same currency. Clients thought the wall was free. It was paid for in tenure, and the invoice was the notice period.

Mark Ritson made the case in Adweek last week that exclusivity is an ancient custom that no longer serves anyone, and most of his evidence is right. Google and Meta hold both colas' plans inside one algorithm. Amazon sells retail media to both. McKinsey, BCG and Bain advise rivals behind ethical walls. In Japan, Dentsu has served Toyota and Honda for decades, and Coca-Cola's own review carves Japan out for Dentsu, which also serves Suntory. The norm, as he says, is historical, not strategic.

But he stops at "abolish it," and that is the wrong stop. You cannot abolish a thing that was never really there. The binary existed on the client side, in the expectation. On the agency side, there was only ever the ladder. What needs to go is not a practice. It is a word, and the question the word produces.

And the counter-case needs saying from inside the room, because the top rung was not always wrong. The fear Al Silk named in his 2012 Harvard study of the exclusivity norm, "divided loyalty," that the agency's best people and best ideas drift to the rival, was never paranoid. For the idea, it is real. A creative team that has cracked your category will crack it again for someone else, and no permission setting stops a human being from carrying a thought down a corridor. For the idea, the wall around the people earns its rung.

For the pipe, it never did.

The first is real: any arrangement in which one client's spend, data or information helps another's. Your identity stack matched against a publisher's, your audiences, your bid logic, your launch date. That conflict can hurt you, and it is bounded, gateable and auditable. Partitioned data, permissions, a log of who touched what and when. A wall around the pipe can be inspected. A wall around a person cannot.

The second is perceived: the need to be the client they think about first. Not a weakness. Andrew Kraft wrote last week that a shared holding company reads as a headline to a board, and the headline is thin. He is right. Perceived conflict is a governance fact.

The error of thirty years was serving both needs with one blunt word, and letting the feeling set the price of the fact.

In The Free Half I wrote that "the margin lived in the mystery, not in the value." Exclusivity is the second thing the agencies gave away in that trade. First the idea, priced at nothing to anchor the media. Then the wall, priced at nothing because scale was assumed to be the premium.

Scale was the premium when a scale account paid for the building. It is not the premium when the building is a platform that serves ten accounts at near-zero marginal cost. Publicis increasingly looks like a technology company, and a technology company's best client does not buy it anything the next client doesn't. The clause outlived the economics that funded it.

Which brings the arithmetic. Count the global-footprint media partners a top-twenty advertiser can actually use. Publicis. Omnicom, with IPG folded in. WPP, mid-turnaround. Dentsu in a set of markets that matter. Three, honestly, and the third is conditional. Now count the category leaders who each expect the whole shop: two colas, three brewers, five carmakers, the telcos, the banks, the luxury houses. Category exclusivity was a rationing device when agencies were plentiful. With three suppliers, it rations the brands. Coca-Cola's global review became a review of one in an afternoon. That is not a Publicis story. It is what the top rung does to a category in a consolidated market, and it will happen again, in beer, in autos, in banking, until the word goes.

In most categories there are now more great brands than there are great agencies. Consolidation in marketing services will only continue. The word will have to give.

The question in that cell should never have been "list your conflicts." It should have been "show us how you manage them." One question invites a list that a nervous reader turns into a verdict. The other invites the ladder, the permissions, the log, and a grown-up conversation about which wall the client actually needs and what they are willing to pay for it in fees, or in tenure, or both.

The industry grows up when the form changes. Not when the holding companies get bigger, or the platforms get smarter, but when the buyer asks the right question.

Exclusivity was the vow. Conflict management was the marriage.

If your agency could show you exactly what sits between your team and your competitor's, the people, the data, the log, would you still need the word, or just the feeling?

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 17 September 2026 in the Brandflow newsletter on LinkedIn.