Rinse and repeat? How the new agency business model will accelerate client in-housing

Opinion

If the legacy extractive media agency business model is dying despite the life-support system of non-transparent revenues, what happens next? asks Nick Manning.

We’ve rinsed our clients, we’ve rinsed the media owners, and we’ve rinsed our people”.

These were the words of a departing senior executive at one of the Holding Companies’ network media agencies about two years ago, describing the pressures of an extractive media business model that has been at full stretch for years.

The model takes advantage of the agency’s role in handling the client’s money as it passes through to media vendors (the ‘principal’ system). As supply chains have proliferated, more money is sliced off, and network agencies get a bigger share in multiple ways.

The group media agencies have long subsidised other business units in the Holding Companies, but now their own fees have also declined, leading them to take bigger slices to compensate for the wholesale decline.

One example: network agencies increasingly outsource online activation to third parties in low-cost jurisdictions. This has six advantages:

• The agencies earn a fee directly from their client
• They mark up the media buying rates back to their clients
• They earn a rebate of up to 40% of volume from the third party
• They can earn a performance-based bonus from the client for ‘outcomes’ generated and validated only by the third party
• They can reduce their own headcount and associated costs
• This is all almost entirely unauditable, and so everything is retained

All of this is much more profitable than the legacy model of having to pitch and win business legitimately, often at ridiculously low terms that are themselves predicated on getting lots of ‘other media income’.

In the current market, pitch competition is insane, so hidden revenue has to be buried deeper.

The effect of this is writ large in two current US court cases; Foster vs WPP describes how ‘principal’ and ‘proprietary’ media arguably kept WPP afloat and was hyper-profitable because its clients declined to participate.

With so much free inventory available, GroupM/WPP Media could use it as it wished, including potentially recharging clients at full cost.

This made it so profitable that WPP’s legal and finance teams had to cap the amount of money being made, not because of advertiser concern but because they feared their own statutory auditors would question the legitimacy of the revenue recognition.

Separately, a class action lawsuit brought by disgruntled WPP shareholders describes how GroupM/WPP Media struggled to both restructure itself and keep up with Publicis, especially on data capability. 

The alleged gap between reality and what WPP told the market underpins the lawsuit, which I examine in Part Two of this article next week.

As an advisor to Richard Foster’s legal team, I have a vested interest in his case, but no involvement in the class action.

Double trouble

The existing extractive model has helped some groups more than others, but, Houston, it creates a double problem for all of them.

Firstly, despite the extraordinary infusion of non-client media income, the Holding Companies’ aggregated profits grew by only a compound 2.3% per annum over the last five years. 

In WPP’s 2024 preliminary results, they reported that “GroupM…grew by 2.7% in 2024…(but) was offset by a 3.9% decline at other Global Integrated Agencies”, 

It can be argued that the ship hit the rocks in 2025, when GroupM/WPP Media’s revenues actually declined, despite the egregious revenues from ‘principal’ and ‘proprietary’ media and the bundled data ‘products’.

The second problem is that these schemes incinerated client trust. The anonymised ANA Media Transparency study blew the lid off, but the Foster vs. WPP papers contain real data and documents that are even more revealing.

The court case papers have exposed several myths about ‘principal’ and ‘proprietary’ media, and advertisers are watching.

For example, the media agencies do not buy media in advance despite what nearly everyone says. It’s what the agencies want you to think, as it supposedly justifies the ‘risk’ premium through media cost arbitrage.

‘Proprietary’ media is another myth, merely a repackaging of client spend.

Perhaps more pertinently, the furore in the Foster vs WPP case surrounding Sony will no doubt have been registered by audit and risk committees worldwide, especially given the vast sums involved and the life sentence handed to a GroupM executive. 

The bigger picture is that the legacy network media agency business model is exhausted, clients are saying “no, thanks,” and trust has been irrevocably eroded.

Specifically, advertisers have resisted ‘principal’ and ‘proprietary’ media, and the extractive nature of the network agency business model more generally. For clients, the juice has not been worth the squeeze, but the agency lemon has been bled dry.

Meanwhile, contracts and compliance audits have tightened, and client-facing agencies have found it harder to get clients to ‘opt in’, especially in the US, where an ‘agency’ culture persists, and the ‘principal’ concept doesn’t fly.

The industry narrative denies all of this, of course, shaped by the Holding Companies and aided and abetted by Forrester. Its report on the adoption of ‘principal’ media seems to be contradicted by all the evidence.

One of its executives even recently said that “media arbitrage may have a reputation and a number of opponents, but it works, and it’s growing”.

Unfortunately, both the reputation problem and the opposition come from some of the world’s leading advertisers. 

So, ‘other media income’ is plentiful, but the lifeboat itself is holed below the waterline, and yet the captains occupy the corner offices. 

So, if the legacy extractive media agency business model is dying despite the life-support system of non-transparent revenues, what happens next?

More of the same on steroids?

The Holding Companies are transitioning to Operating Companies (OpCos) and creating new ‘black boxes’ that bundle together the whole content, production, distribution, commerce, audience targeting, data, analytics and ‘outcomes’ delivery process.

The new business model behind these platforms will almost certainly apply extractive techniques across the entire advertising life cycle, not just media. More control means more margin.

Anything that can be marked up probably will be, especially if bought in, and the ‘proprietary’ media approach will be applicable to all media inventory, not just the low percentages of budget we see today.

In fact, the ‘proprietary’ model will be the model, and it bundles up the full set of client requirements into a supposedly neat package where ‘outcomes’ are guaranteed (as long as the word ‘outcome’ is self-defined). 

Fees can be zero when the margins are embedded, giving the appearance that it’s all ‘working’ money now.

Meanwhile, the entire end-to-end process will be increasingly automated, with ‘agentic’ technology that eliminates the need to target people and employ them.

The OpCos’ strategy no longer rests on ‘big’ advertising as an engine for expansive business growth, something that can shift perceptions at scale and at speed, open up new markets, build brands and justify prices and thus drive profit growth. 

Instead, we’re moving wholesale to ‘small’ advertising, and from ‘showmanship’ to ‘salesmanship’ (to quote Orlando Wood) while losing our prior focus on the public we served, the brands we built, the human capital we nurtured and the media channels we supported.

‘Creative’ is now ‘content’, with thousands of variants to fill the spaces for many billions of impressions on an increasing number of channels. If content used to be King, distribution is now Emperor.

Advertising is now a series of transactions, mostly micro, with audiences reached in singles, not sixes or fours, with self-reported metrics, or none at all.

There is little accountability or transparency, and that’s how the agencies want to keep it, despite protestations to the contrary.

And, perhaps most harmful of all, smart advertising thinking founded on consumer truths is squeezed out in favour of data analytics, largely self-reported.

Will this ‘work’? Probably not for advertisers.

We already have two clear case studies of how this plays out: the Open Web programmatic trading system and the ‘principal’ media models.

Both enrich the agency and ad tech communities to the detriment of advertisers and media owners. We can include the public in this, given how programmatic trading contributes to terrible user experiences.

All of this has been proven time and time again, with evidence galore, including the latest court papers. This is not new news.

Both programmatic and ‘proprietary’ models rely on one key dimension: it is virtually impossible to tell whether the morass of advertising being paid for even appears, reaches an audience (real or not), or gains any real attention, engagement, or impact. 

The transacting parties get paid anyway, increasingly in ways invisible to the advertiser, and by owning a big chunk of the data layer (e.g., via Infosum and LiveRamp), the OpCos can control more of the budget and the data.

Two key questions: is this really what advertisers want and need, and who benefits most?

If clients increasingly view advertising as an amplified form of direct response, addressing the ‘in market’ buyer on a one-to-one basis but supposedly at scale, then it might make sense for some advertisers. 

For most advertisers and brands, the transformative effects of collective targeting with high standards of accountability and transparency won’t die quietly. They may be less immediately measurable, but the cash register doesn’t lie.

Much has been made of PepsiCo’s move to Publicis, but Michael Farmer’s comments via ADOTAT are apt; PepsiCo has stagnated for years, so will improved data and analytics revive its fortunes, especially if it goes ‘small’?

As for who wins financially, the smart money should be on the OpCos. They can apply the extractive model without asking clients to acquiesce, or ‘opt in’, solving that problem.

In the ‘black box’ future, everything will be far more opaque, and the contracts will be completely redesigned to keep it that way. 

What should advertisers do now, and how will this change the advertising industry?

The answer may be a cliché, but they should take back control of their fortunes and apply the kind of rigour and discipline to marketing that they accord to other capital-intensive activities.

Marketing remains the key growth driver but remains undervalued and under-scrutinised within companies.

The incessant talk of marketing needing to prove its worth and take greater accountability for profit should lead to a far stronger focus on true business results. The ‘closed loop’, ‘black box’ and ‘outcomes’ trend within the agency supply chain does not provide this. The talk of ‘outcomes’ is mostly hogwash.

In the real world, advertisers need to know what their marketing investments produce for them individually and in totality, with measurement that shows the true effect.

They need to know what is happening to their data, that they are not exposed to cybercrime, privacy scandals, or leakage, and to be wary of LLM predation and the resulting IP risk and deceptive practices. These are already legion.

They need to fully track their data and their money and quantify their efforts and investments. 

In short, they may need to move much more work in-house to do this with a high degree of confidence, security, and efficiency. They will bring in-house strategic marketing thinking, data, analytics, and parts of the activation process.

They will work directly with Amazon, Meta, Google, TikTok and some Retail Media partners and streamers.

What about the OpCos? Their role will probably be limited to parts of the activation process, but they will have to be far more open about their operating model.

Advertisers may conclude they have been rinsed enough by partners they thought they had appointed to act in their best interests, using their data and money to protect their brands. They may not want to get fooled again.

The OpCo strategy could lead advertisers to decide they don’t want to outsource the husbandry of vast sums of money and data to outside parties within ‘closed loop’ systems they can’t track.

This may also stimulate the independent agency sector for those clients who don’t need the industrial-strength international reach and capabilities the OpCos purport to provide.

It may also spark an outbreak of independent data and tech providers with whiter boxes who charge transparently.

It will spawn a new generation of independent companies who can make sense of marketing through smart analytics.

In short, advertisers may choose not to be rinsed again. They will need to take back control and surround themselves with advisors who can and do provide impartial and well-informed navigation.

Agencies are supposed to help by navigating choppy waters and clarifying and simplifying; some no longer see this as their primary role.

This is a prediction, and the jury is still out on a market in the throes of moving from a highly dysfunctional system to a new one where AI and ‘agentic’ targeting and trading are more advanced, and probably even worse for advertisers.

In Part Two, the pain of a market in transition is examined in excruciating detail via the newly released court papers in the WPP class action lawsuit.

They describe the troubled environment within WPP from 2024 to 2025, and the human cost of maintaining pretence in the face of everyday reality. It’s a lesson for everyone.

Nick Manning is the co-founder of Manning Gottlieb Media (now MG OMD) and was chief strategy officer at Ebiquity for over a decade. He now owns a mentoring business, Encyclomedia, which offers strategic advice to companies in the media and advertising industries, and is the non-executive chair of Media Marketing Compliance.