What you need to know:

• Principal-based media – where agency groups buy media and then resell it to brands, often via non-disclosed, non-auditable models for a handsome mark-up – are on the rise and coming to a head.
• Publishers warn overcharging for their inventory – blended with cheaper alternatives – is distorting the market, creating perceptions of ineffective investments and pushing them off the plan.
• Consultants say agency groups risk losing advisory roles and regressing to basic media buyer status due to distortions and conflicts arising from their incentives.
• Holdcos are stepping up principal activities – and some see nothing wrong with the myriad principal models, provided clients understand their contracts.
• “A smart client, I would argue, realises the agency is able to provide the overall balance of services because of what [a principal model] affords – because no client will properly pay full freight for [media],” per ex-GroupM global business intelligence chief turned Madison and Wall Principal, Brian Wieser.
• “It's flawed, but it's better than the next best alternative.”
• Others strongly disagree: “I don't think principal-based trading is in the best interests of anyone,” per one ex-holdco global exec. “It's not in the best interest of clients, it's not the best interest of agencies – and I think people who support the model don't fundamentally get the flaws with it.”
• Dan Johns, a former agency boss at Ikon and Havas Media and one-time COO at IPG Mediabrands turned consultant at Tumbleturn Marketing Advisory, thinks even sophisticated marketing operations often don’t understand enough about the principal contracts they are signing.
• He thinks principal models are distorting the market and advises clients to ask a fundamental question: “Are they okay in a scenario where an agency has a vested interest in making a recommendation into a particular channel?”
• Johns sees a growing role for impartial media advice – and says creative agencies are poised to fill the void.

Former GroupM global business intelligence chief Brian Wieser said at the start of the year that marketers “kind of accept, if not sometimes prefer,” principal media models. It was one of Mi3’s highest read stories of 2024.

As the year comes to a close, Wieser stands by it – and other holdco bosses, most recently IPG’s, say clients “accept and even embrace principal buying”. So the signals are clear.

“To say clients ‘like’ it is an overstatement,” per Wieser, who founded Madison and Wall in early 2023. “What I'm saying is it's flawed, but it's better than the next best alternative. All things in life, in business, it's rare that we get what we want. We look for the least bad alternative relative to other choices available.”

Either way, all the signs suggest principal trading is on the rise, globally and locally.

“It definitely is,” says Wieser.

But not all arbitrage is equal.

Definitions, constructs, MMM shortfalls

“Some of this comes down to definitions … There's so many different flavours of deals, you could probably come up with a list of 100 different kinds of transactions, then you layer on top of that transparent and non-transparent. Then you layer on top of that, ‘best efforts’, as I characterise it – and ‘best efforts’ deals are everywhere,” says Wieser.

His ‘best effort’ definition is where the agency “makes it known to the seller it will be worth their while to provide a more favourable transaction, but they can't commit to it contractually”. According to Wieser, “that’s just good commercial trading”.

What's different now versus 2016 when agencies exploiting ambiguous contracts caused a transparency blow-up is that “contracts are much more buttoned up,” per Wieser.

“Layer on top of that the fact that you have ‘complicit client syndrome’, where the client knows they want the overall service offering they're getting from the agency,” he adds.

“In order to get that overall balance of what they want, they know they need to let the agency do what will allow them to show a lower weighted cost [for media].  Because then procurement will be happy – which makes the client a hero and gives them more latitude to do the thing they need to do.”

True, but some would argue that focusing on discounted media and cross-subsidy models negates efficiency.

“I don't think so,” says Wieser. “Anyone who says they can tell you what the efficacy of media is to start with is either ignorant or lying,”

Isn’t that what market mix modelling – or MMM – is supposed to be good for?

“You can't tell me that market mix models capture everything. There's a huge amount of imperfection in all of this to begin with … There is a role [for MMM], but I’m just saying you don't want to overestimate its precision and perfection.”

Either way, “media effectiveness is not what this is about”, per Wieser.

“If it were about effectiveness in media, everybody would be using Wieden and Kennedy … and optimising not the price per unit of spend, but the overall campaigns.

“That is not the way the world works. It's about satisfying a marketing goal within the constructs that companies have set up for themselves.”

'The smart money'

Wieser suggests marketers knowingly committing to arbitrage models may be the smart ones.

“If you're the client that's providing better overall economics for your agency, are you going to get better service? Absolutely. Are you going to get better strategic ideas? Absolutely – as long as you're a good client and know what you're doing. That part is missed in all of this.”

But are these deals not distorting the market?

“They are distorting the market, for sure,” Wieser acknowledges. But he thinks some of the claims about how much money the big agency groups are making from arbitrage models are overblown.

“We don't normally know, publicly at least, what the margins are for the media agencies. We can roughly estimate – I certainly know what GroupM’s was – and there's enough gossip around what these numbers are to roughly triangulate around what the margins of the media agencies should be. So we know it is a growth engine of the business.”

But he said some of the claims being made on just how much upside holdcos are banking from principal models are “just fantastical”. It’s a driver, per Wieser, but it’s not the driver.

Is it not the fastest-growing part of at least some holdco’s businesses?

“That's really hard to say. I couldn't prove that,” says Wieser. “I couldn't disprove it.”

Conflicted but better?

Wieser is applying a version of what principal media enables to his own revenue model at Madison and Wall, which he launched in February 2023 after leaving GroupM.

“I've learned so much more about running businesses in the last year and a half than I ever did in the rest of my career,” he told Mi3.

“I found that some clients had budgets for consulting, advisory, but they had no budget for research. Some had a data budget. Some have speaking fee budgets. So I figured, what if I bundle these things together? Initially I had started by selling my time, and I just gave away the research with it. But then as last year evolved into this year, I started realising I'm going to run out of time if I keep doing that. So I started to say, ‘if you spend more than 2x the cost of a subscription on services, you get the data and the research for free.’

“My goal actually is trying to price like the way like [Publicis data unit] Epsilon goes to market – you can't buy the data independently of the service.”

Wieser draws parallels between principal-based models and the separation of investment banking and research, instigated in 2003 by then New York Attorney General, Eliot Spitzer. The upshot meant investment bankers could no longer use investment banking fee revenue to compensate research analysts. Prior to that, banks were effectively underwriting analysts.

“It was conflicted, yes, but the resources that went into it were fantastic, and it meant that the overall quality of the work was great,” argues Wieser, who at the time was a Deutsche Bank analyst. After the separation, “The quality of research went way down because the subsidy for ideas went away. People have a really hard time justifying paying for ideas.”

In other words, “the old world that was conflicted produced better research, and if you understood the conflicts and understood how to consume it, a better offering for everyone.”

And there lies the rub with principal media models. Despite everything that’s gone on over the last few years, including ACCC-led inquiries that essentially told clients to wise up, plenty of anecdotal evidence suggests many marketers – certainly in Australia – don’t actually know what they are contractually agreeing to.

Wieser gives that short shrift. He quotes Machiavelli:

“A prince who is not himself wise cannot wisely be advised.”

“They should. That's their fault. There's no excuse at this point in time for clients to not understand this stuff and not invest in people and processes to understand it.”

But he reiterates the flipside – that “complicit clients who do understand this and recognise the overall benefits they're getting outweigh the costs and the downsides.”

In the US and Europe, Wieser’s observation is that “complicit clients are way more common than not. They understand more or less [what they are getting into]”, though they may not know “everything they could know”.

He can’t say whether the same applies in Australia. But as long as contracts are buttoned down on specifics, “that is perfectly clean”.

Regardless of region, Wieser sees a simple bottom line on principal media:

"A smart client, I would argue, realises the agency is able to provide the overall balance of services because of what this affords – because no client will properly pay full freight for it.”

‘Fundamentally flawed’

One senior exec – formerly with a global remit at a holding company – disagrees.

“I don't think principal-based trading is in the best interests of anyone. It's not in the best interest of clients, it's not the best interest of agencies – and I think people who support the model don't fundamentally get the flaws with it,” he suggests.

“Holding companies once aggregated and negotiated volume and got discounts for clients. They were buying what was in the client's best interest. The bit that everyone misses on the principal-based trading model, and where it all falls apart, is holding companies are now negotiating what's in the best interest for themselves.”

He thinks there could be legal implications – i.e. acting as a ‘faithless agent’ – though it’s more the principle of principal media that is problematic.

The exec provides a theoretical example.

“Let's say you as a client want to buy channel Nine, and channel Nine won't enter into a ‘great deal’ with the holding company, but [hypothetically] channel Seven will. So even if channel Nine has your right audience in the right place, the holding company will just put channel Seven on your plan and not tell you otherwise. And so that's where principal-based trading absolutely kills you. You end up getting media that you don't want. It's like your dietitian running a McDonald's restaurant and saying, ‘Trust me, I'm going to tell you what to eat’ – and they just keep giving you McDonald's.

“You ask yourself, 'is this right? Is this good for me?' They tell you, ‘Yes, it’s in your own best interest to eat more McDonald’s – and it’s a good price'. You think, ‘this doesn’t seem right, but they are a dietitian…’”

Does it matter if the media is delivering a good outcome and a return on the investment?

“As a marketer and as a client, you've got no basis to actually fact check the accuracy of alternatives. So yes, there's an ROI there, but you don't know if you were going to get a better ROI. You don't know if you're going to get a better deal on channel Nine or better ROI on channel Seven, it's just not fundamentally there,” per the exec.

“So the removal of choice just means you have to accept what you're given – and keep rolling with it.”

Given latest peer-reviewed marketing science highlights the increasing limitations of reach-based planning, more opaque principal models could pose further problems for marketers seeking an edge on competitors by better exploiting category and channel nuances versus the “really mediocre outcomes” observed across 1,000-plus reach-based campaigns and 1 million customer journeys.

In short, if brands can’t accurately map where their money has been spent to sales outcomes, how can they optimise to growth?

Big tech, bigger problem

The ex-holdco global exec thinks increasingly cosy relationships between big tech companies and holding companies actually pose a bigger problem than arbitrage via principal-based media.

“[Big tech platforms are] essentially incentivising holding companies on the adoption of certain products or services, and they're tied to penetration of those products [slash] services across the client set.”

“So they're becoming resellers in some ways, but non-disclosed resellers.”

Versus media arbitrage, “I think the incentives on the tech side of things are greater”, per the exec.

“The tentacles of the tech companies are in so many parts of the pie. There's a significant amount of distortion in the marketplace, in the alliances and questionable partnerships in place.”

Can it be fixed?

“It's either [going to require] competition reform or regulatory intervention,” he suggests, because industry is unable to extract itself from the model it has constructed.

“Holding companies have built a prison that they can't get out of. They're relying on principal-based trading deals, relying on payments back from media companies to fund their business model … I don't know how publicly listed holding companies can actually adjust to the changes in market dynamics that are inevitably going to come.”

Hence antitrust action in the US could eventually have major ramifications for agency groups now reliant on big tech incentives.

“Google's in the spotlight, Amazon's in the spotlight. Once those companies get investigated – and if they get broken up – you can see those deals evaporate.”

Australia view: Pressure building

Principal-based trading models are rising sharply in Australia.

“It's a growing trend within the market. It definitely is becoming more prevalent,” says Tumbleturn Marketing Advisory partner, Dan Johns, a former agency boss at Ikon and Havas Media and one-time COO at IPG Mediabrands.

“It's becoming a more obvious part of contracts – and where principal media is in place, we're seeing growth in terms of investment into principal media over time.”

A growing number of independents are also now buying media and arbitraging – though some are holding out.

“I can't categorically say that every single group is [using principal-based models], but my assumption is that all of the major groups with volume are trading in that way. I think it's a standard part of the way in which a large group will operate,” said Johns.

“You've got the counter-trend of smaller agencies putting arguments against [principal models] into the market.

“The other dynamic is media owners … and it feels like that is about to surface; it feels like that is where the pain hits the most,” he added.

The problem is, if one media owner is discounting via principal trades, it’s hard for others to hold out, especially those under heightened financial pressure – though some are trying to build more direct relationships to counter margin hits.

“We definitely hear the frustration at a senior level from that side of the market,” says Johns. But on the flip side, “it can provide a convenient way for them to have some level of surety around the revenue.”

Distortions, 100% mark-ups

Johns said there is no doubt that principal models are causing market distortions.

“That is definitely the case … wherever there's a revenue stream for a channel, it distorts the outcome.”

He agreed with the ex-global holdco exec that marketers can’t compare like-for-like what would have happened without an agency incentive to pick a specific channel or publisher within that channel.

“The whole point around principal [media] is that there's a better savings outcome – you can buy cheaper. But what are the benchmarks to that? It isn't necessarily clear what the comparison is,” per Johns.

“We were talking to a head of sales from one of the networks, and his major concern was the degree of inflation that agencies were putting on to BVOD actually prices it out of being effective.”

That’s the same issue flagged by Mi3 last month: An advertiser assumed they were buying broadcaster BVOD via a principal media deal – but the CPM rate was double what broadcasters are selling at and was loaded with cheaper connected TV (CTV) ad slots.

Some holdcos are packaging this blend as ‘TVOD’ rather than BVOD – but the upshot is that advertisers are assuming BVOD is expensive and doesn’t deliver. The reality is that they are paying way over the odds for a diluted blend while the broadcaster gets squeezed. Or as one holdco exec put it, “We’re killing the industry with short-term greed.”

Even without BVOD-CTV blending, Johns said those kind of distortions are problematic.

“A major concern for publishers is the additional margin put on their inventory by agencies – they lose complete control over their pricing approach. Inventory that might leave the business at $40cpm can appear on a client plan at double this.”

Do marketers care?

Whereas Madison and Wall’s Wieser thinks marketers are complicit – actively opting for principal deals to pay for other services they need while keeping procurement happy – Johns is less sure.

“On all these transparency issues, clients aren’t [fully] aware – and nobody likes to admit that. There is a mild understanding of what is going on without really understanding… Every client is different, but even some of those with quite large marketing departments, with big procurement departments with dedicated media teams, don’t necessarily have an in-depth understanding.”

Johns says principal models do have a “stronger value exchange” than previous grey areas like rebates or non-disclosed programmatic – but those behaviours are not necessarily high benchmarks.

“The agencies are good at packaging principal media to show a reduction of fees can be funded through [that mechanic]. So on the surface, the client can see a strong value exchange without really understanding the fact that they can be opting out of any level of audit right or transparency around the way in which that media is traded.”

Johns said attitudes can change quite quickly when marketers are told what to look for.

“When we work with clients, we flag it and [in cases] contracts are changed.”

While all clients are different, Johns said marketers face the same fundamental choice.

“The root of the issue is around an agency's ability to be able to give impartial advice. That's a crucial question that clients need to be asking: Are they okay in a scenario where an agency has a vested interest in making a recommendation into a particular channel?”

Intermediaries disintermediated?

If more money continues to flow into principal models, Johns said media agencies could effectively end up as media owners.

“That seems like a fundamental shift. We’ve seen big changes in industry over the last 30 years – this could be the catalyst to another big one.”

Such a shift could potentially regress media agencies from partners back to just media buyers – which is problematic given most buying is already automated.

“That is the ultimate consequence … it goes all the way back to how it was 30 years ago and that [strategic] void is filled by someone else,” per Johns.

He said creative agencies are already positioning to handle “connections thinking” while eschewing full service approaches, as doing the buying as well as the strategy would create the same conflicts.

“There's a re-emergence of strategic capability … We went through an era of independent agencies popping up and making a strong play around impartial advice – the Naked approach,” said Johns. “It feels like that there is an even more significant need for that level of capability today.”

The question is whether marketers and procurement are willing to pay for strategic advice. In the end, Naked couldn’t make that model work – and the rise of principal models suggests otherwise.

“I think there will come a point when everyone realises the consequences of not being strategic and not having that resource,” said Johns. “You end up in a world that on the surface might have some advantage to it in the short-term – but ultimately won't drive the business forward.”

Are there any signs from Tumbleturn's client conversations that tipping point will arrive any time soon?

“Yes. Every client is different and will have a different level of tolerance and a different perspective whether they care or not … but certainly conversations with some clients progress very quickly from unawareness of non-disclosed and principal media techniques through to reconsidering the existing relationship with the agency … and then starting to think about how things could play out differently in the future,” said Johns.

“Definitely.”