The advertising industry has always preferred its own fairytales.

Every agency claims to compete on brilliance, integrity, and strategy. The language is calm, practiced, and delivered with just enough confidence to make skepticism feel impolite. Transparency is invoked often, usually on stage, occasionally in decks, and almost never in ways that meaningfully alter the economics underneath.

Clients participate because it is comforting to believe that a “trusted partner” is engaged in a principled fight against inefficiency. Executives gesture toward slides about openness as if repetition alone neutralizes conflicts of interest. It is a familiar performance. The audience is large. The script rarely changes.

Judy Shapiro, who spent more than a decade inside a major agency before moving client-side, has been unsentimental about where this leads. Her view is blunt. The agency business model was built for a different era, and it has been deteriorating for ten to fifteen years while leadership postponed a reckoning. The decline was not sudden. It was managed, rationalized, and softened through restructurings, mergers, and a constantly refreshed vocabulary of transformation.

That context matters, because it reframes the industry’s most persistent debate.

This is not a sector struggling to fix a broken system.
This is a sector protecting a system that still works extremely well for those inside it.

Behind the polished decks and glass walls, the machinery operates exactly as designed. Agencies no longer survive primarily on intellectual advantage or strategic insight. They survive on engineered margin, structural opacity, and commercial complexity, all calibrated to exploit how difficult modern media economics are for clients to fully interpret.

A senior operator inside a major holding company described it plainly on background. Principal trading, media curation, and fee arbitrage did not disappear. They migrated into instruments that look like execution rather than compensation. Functionally, they are all substitutes for fees. The defining feature is not superior performance, but deniability, depending on how the structure is framed.

Nick Manning, co-founder of Manning Gottlieb Media, former CEO of OMD UK, longtime Ebiquity executive, and author of How Principal Media Is a Game Where Everyone Loses, reaches the same conclusion from a different angle. After decades running agencies and nearly twenty years analyzing them independently, Manning argues that principal media is no longer peripheral. It is a core feature of the holding-company profit model, relied upon to inflate margins, subsidize declining divisions, and offset the structural compression caused by procurement pressure and automation.

If transparency is discussed constantly but rarely delivered, the reason is not subtle.

The system works.
Transparency is not missing.
Transparency is threatening.

What follows is not the teardown yet.
It is the map.

At the center of the modern agency P&L sits arbitrage.

It is the practice of buying media at one price and reselling it at another, reframed over the past decade as innovation. Enhanced access. Premium supply. Value unlocks. Each phrase exists to soften what is, in operational terms, media trading.

Once the language is stripped away, the behavior is clear.

Media is being flipped.

This is not hidden, and it is not fringe. Holding companies reference principal trading in investor communications as a core economic lever. It appears as margin diversification, not creative differentiation. Earnings calls do not linger on strategy. They focus on spread.

The holding-company operator was explicit about why principal positions dominate. They carry the most economic weight because they scale fastest and sit closest to spend flow. Rebates are largely linear and increasingly irrelevant. Principal positions compound. That compounding is what makes them powerful.

It is also what makes them dangerous.

Manning reinforces this point with financial clarity. In his analysis, principal media generates super-profits, often far exceeding disclosed fees, and has become essential to holding-company earnings stability. Media, in effect, now subsidizes the rest of the organization, including investment in proprietary tech and AI initiatives.

The danger is structural. Principal trading must survive one of two outcomes. It must either prove incremental performance under scrutiny, or it must quietly reclassify itself as disclosed execution. Tech spreads endure because they can be reframed as infrastructure costs. Principal trading cannot. If it fails to demonstrate incrementality, it is no longer value creation. It is rent extraction.

Michael Farmer has long explained how the industry arrived here. Farmer is the Chairman & CEO of Farmer & Company LLC, a role he has held since 1990, advising agencies and advertisers on scope-of-work design, workload measurement, and remuneration reform. Through his proprietary ScopeMetrics® system and decades of executive training, Farmer has documented how procurement-led fee compression severed compensation from actual labor and complexity.

As fees were pushed down and scopes expanded, agencies absorbed the pressure until they couldn’t. Creative agencies had limited room to maneuver. Media agencies did not. Principal trading became economic triage, stabilizing revenue without fixing the mismatch between labor, scope, and compensation.

Internally, performance volatility might generate debate. KPI movement might prompt discussion. But the operator was candid about the hierarchy. Principal positions scale faster than expertise, and margin protection wins every time.

Graeme Blake of Bluetui® captures the resulting conflict succinctly. When agencies frame principal trading or curation as value creation, that value often derives from prioritizing inventory they already own or control. The truth shows up in optimization behavior. Real performance creates uncomfortable outcomes: sometimes spending less on owned supply, sometimes recommending against internal margin, sometimes shifting volume away from house inventory mid-flight.

When that never happens, the incentive structure is visible.
The feedback loop resembles grading one’s own homework.
Independence becomes theoretical.

The second pillar of agency economics is quieter but equally durable: fee stacking.

Agencies learned long ago that scrutiny follows large numbers. It rarely follows small ones. The solution was fragmentation.

Technology fees.
Data fees.
Measurement fees.
Optimization fees.

Each looks modest in isolation. Together, they transform economics.

A handful of fees irritates.
A dozen forms an ecosystem.
Enough of them becomes the business model.

The holding-company operator noted that agencies that have “won” in recent years are not necessarily the ones with better performance, but the ones that became more creative with pricing while maintaining plausible deniability. Agencies that took strong stances on transparency and appropriateness were often punished commercially. The industry says it wants clarity. It does not reliably reward it.

Manning documents the same pattern historically. As older non-transparent revenue schemes were exposed, principal media emerged as the “acceptable face” of opacity. Cash rebates gave way to inventory deals. Bundling followed. Media and service costs blurred. Opacity migrated rather than disappeared.

By the time fee fragmentation is complete, a client that negotiated a five-percent headline fee may be paying an effective take rate several times higher. Because charges are distributed across entities, scopes, and classifications, no single line item triggers intervention.

Farmer traces this to a structural failure. Agencies never built durable systems to align workload with compensation at scale. When procurement demanded cuts, agencies complied visibly and recovered margin elsewhere.

Visible fees declined.
Underlying economics did not.

Shapiro has been direct about the cost. Headcount reductions followed. Institutional knowledge exited. Capability eroded. What remained was a structure increasingly dependent on opacity to function.

Complexity was not accidental.
Complexity became cover.

If creativity and strategy no longer define the holding-company advantage, something else does.

It is the ability to route spend through structures that resist inspection.

The configuration is familiar. The contracted agency. An affiliated trading desk. An investment unit operating beyond contractual scope. Partners that appear independent while sharing corporate parentage.

When clients ask about transparency, agencies point to audit rights. Those rights often exist. They also tend to end precisely where the most consequential economics begin.

Both Manning and the holding-company operator agree this is not an ethical failure first. It is a financial one. Internal P&Ls are not designed for zero-opacity flows. If a client asked the right questions and fully understood what they were looking at, finance breaks first. Revenue recognition, margin attribution, and intercompany transfers unravel. Media follows, when optimization narratives collide with disclosed incentives. Legal steps in last to slow the blast radius.

Opacity is not an oversight.
Opacity is the product.

Remove it and spreads narrow. Rebates disappear. Principal trading becomes harder to justify. Enterprise value, in many cases, is tied to margins that do not survive transparency.

This is where pressure accelerates.

AI has not collapsed labor, but many agencies are behaving as if it has. Operationally, automation has helped. It has not replaced the people required to mine, synthesize, interpret, plan, and optimize media in market. Not even close.

What AI has exposed is that labor was always undervalued. Personnel are increasingly treated as decision liabilities rather than specialists accountable to outcomes. Responsibility inside brands has bifurcated between media live and media accountable. Causal attribution has become muddier, not clearer.

Here again, the sources converge. The correct framework is widely understood. Pricing against incremental lift. Auditable systems. Pre-registered hypotheses. Shared measurement frameworks. Fees tied to contribution margin.

The operator’s assessment was blunt. That framework puts agency economics at risk when performance fails. Which is why it remains largely theoretical.

Every few years, the industry announces a renewed commitment to transparency. Each time, it fades.

Not because transparency is impractical.
Because full transparency destabilizes the economic structure it would reveal.

The industry knows the right answer. It also knows how to obfuscate it, appear forward-leaning, and recreate a black-box model under a new name. Good stories sell. Billables speak louder than principles.

Brands often do the opposite of what they say they want. Agencies respond rationally. Efficient reach plus fragile measurement becomes acceptable again, only now it is branded differently.

This column is orientation, not conclusion.

The work ahead examines how principal trading functions in practice, how curated supply absorbs margin, how AI accelerates exposure, and what happens when visibility reaches the financial core.

Many have suspected these dynamics for years.

The difference now is that the math, the technology, and the incentives are converging.

The blindfold is no longer necessary.

Stay Bold, Stay Curious, and Know More than You Did Yesterday.

Michael Farmer is not reacting to AI, procurement, or holding-company chaos.
He’s documenting the consequences.

In his upcoming book, Madison Avenue Revisited, Farmer labels the industry’s decline “murder, not manslaughter.” Not an accident. Not disruption. A long series of deliberate leadership choices that hollowed agencies out while pretending nothing fundamental was broken.

His core argument is simple and brutal: once agencies moved from 15% media commissions to labor-based fees, they needed to understand how work was done, scoped, and measured. They refused. Farmer writes that agencies developed a deep resistance to measurement, convinced it would commoditize creativity. Instead, he notes, it normalized chronic downsizing.

Procurement didn’t destroy agencies. Agencies failed procurement first. Farmer recounts how procurement asked agencies to explain how ads were made and what metrics should be used. Agencies dismissed the request. Procurement responded by cutting fees year after year, assuming agencies must be profitable if they could survive the cuts. Farmer documents the result plainly: capabilities eroded, culture collapsed, and leadership looked away.

Holding companies, in Farmer’s telling, didn’t reverse the damage. They accelerated it. He reports that executives struggled to name any holding-company decisions that materially improved advertising effectiveness. What they could point to were margin targets, headcount reductions, and financial extraction. Farmer uses the word “milked” intentionally.

Principal media, rebates, and non-transparent operations show up in Farmer’s work not as clever innovations, but as economic coping mechanisms. When labor fees stopped working and procurement locked them down, agencies moved margin elsewhere. Farmer lists non-transparent and principal-based media as one of the major crises mishandled by holding-company leadership.

AI doesn’t save this model. Farmer calls it out directly: “shit in, shit out.” AI requires senior judgment, experience, and creative leadership. The same people agencies spent two decades cutting. AI doesn’t fix the business model. It exposes how little of it was ever understood.

What makes Farmer’s book matter now is not that it predicts collapse.
It explains why collapse was inevitable.

The industry didn’t lose its way.
It chose it.

Stay Bold, Stay Curious, and Know More Than You Did Yesterday.

If you think the free version was spicy, the paid section is where the real skeletons come out of the holding-company closet.

Inside ADOTAT+, we don’t just explain why Razorfish keeps getting boxed out by Digitas, Sapient, Burnett, and the Media spine. We unpack the receipts: the pitch detours, the talent flight, the internal politics, the P&L gravity games, and the uncomfortable questions Publicis execs whisper but won’t put on a slide. It’s the part where the agency fairy tales end and the portfolio math begins.

If you want the internal logic, the off-the-record commentary, and the very clear picture of which agencies are part of the future and which ones are being quietly prepared for “redistribution of capability,” that’s ADOTAT+ territory.

Subscribe now. Or stay confused. Your call.

(This link is 50% off fyi, only this link)