The Mirage Metric

A mirage is not a lie. It is a play of light with real conditions producing a false image. Convincing from a distance. Dissolving as you approach. Michael Farmer's creative pricing chart is a mirage. The data is real. The conditions that created it are real. But what people are seeing when they look at it is not what's there.

Let's get closer.

It was when Andrew Tindall called this "the chart of the decade" that I knew I needed to look more closely. In my experience, populist oversimplification of received wisdom with confidence and clarity is not helpful. And the last thing a complicated question needs is the comfort of a clean answer.

Between him and the Wall Street Journal's CMO Today picking it up, it is now actively shaping how CEOs and CFOs think about agency value in real time. The misreading has moved from LinkedIn into boardrooms. Mirages are most dangerous when they influence decisions.

Here is the chart in question.

The chart tracks the price per ScopeMetric Unit, which is Farmer's proprietary measure of creative agency output, where one unit equates roughly to the effort required to produce a standard 30-second TV commercial across 33 years of client data. The headline finding: price per unit has fallen by 75 percent in real terms, from approximately $435,000 in 1990 to approximately $110,000 today. The interpretation being widely shared: agencies have allowed their value to collapse.

That interpretation is what I want to challenge.

Let me start by saying that I have enormous respect for Farmer. His decades of work on scope management from the trenches and not the sidelines have done more to clarify the structural dysfunction in agency-client relationships than almost anything else published in this industry. His books are serious. His data collection is patient and meticulous. He has spent over thirty years building a methodology in a field where most people only visit.

Which is exactly why I'm concerned about what's happening to this chart.

What was originally careful, qualified analysis of scope management data has been amplified into a sweeping indictment of an entire industry's value. Farmer is too rigorous for that conclusion. The narrative has gotten ahead of the evidence, and in a media cycle this fast, that is easy to allow.

Mirages form fastest when conditions are right. And the conditions in this industry, budget pressure, AI disruption, platform disintermediation and talent compression have created an audience that is thirsty enough to believe in water that isn't there. The more pressure an industry is under, the more urgently it needs its fears confirmed. That urgency is what turns a data point into a verdict.

With that said. Let's take three steps closer to the mirage. The correct summary is this: the data is real, the conditions that created it are real, and the image it produces is almost entirely misleading.

Imagine someone handed you a blank version of this chart. X axis: 1990 to 2023. Y axis: price per unit of creative agency work. Then asked you to estimate the line.

If you understood anything about how this industry changed over those 33 years, you would draw it heading steeply downward. Not because agencies destroyed their value. Because the chart could only go one way. You would draw this chart and you would be right.

Take the first step closer, and the crisis image starts to thin as three structural forces guarantee this.

1. The scope explosion. When volume of output per campaign multiplies by 25 to 50 times, the price per unit mathematically compresses, even if total revenue grows substantially. This is arithmetic, not decline. This was inevitable the moment digital channels multiplied the surface area of a campaign beyond anything traditional production models were designed for.

2. The production cost collapse. This is critical to understanding what the chart actually measures. The ScopeMetric captures the total agency fee charged per unit of creative work, the billing rate if you like, not the raw production cost. The digital equivalent of the production required for campaign work today continues to compress, as it should. AI is now accelerating it. Content production time dropped 80%. Social video that previously required an agency team now runs under $100 with the right tools. If production costs fell 80 to 95 percent while price per unit fell 75 percent, agencies may have maintained or improved the underlying margin on each unit. The chart shows you one side of an equation that requires both sides to mean anything.

3. The unit itself changed. Farmer's measure assigns values based on estimated creative man-hours where a 30-second TV origination equals 1.43 ScopeMetric units; a banner ad (whatever that is…) is 1/135th of that. Comparing their prices across 33 years is like measuring the price per phone call in 1990 and 2023 and concluding that the telecoms industry destroyed its value, when total telecoms revenue quadrupled over the same period.

Global advertising expenditure grew from approximately $275 billion in 1990 to approximately $856 billion in 2023. The combined revenue of the Big Four holding companies grew from roughly $7.5 billion to approximately $60 billion today.

The chart was always going to look like this. The question is whether anyone should be surprised by a line that volume expansion, cost deflation, and technological change all but guaranteed.

They shouldn't. First step closer, and already the image is shifting.

The question being widely extracted from this data "how did the industry allow its value to decline?" is the wrong question.

Take the second step closer. The mirage shifts again.

The correct question is this: with price per unit of creative work decreasing, and cost per unit decreasing alongside it, did the industry reduce its cost to supply each unit fast enough, and did it successfully capture value elsewhere in the portfolio? (I get it, not as punchy, but this is the real question)

For the large holding companies, this line going down was not an accident. It was a portfolio choice.

The holding company model that dominated agency growth for three decades was built on a specific commercial logic: creative services would be priced at lower and lower margins in order to secure and justify media relationships that operated at substantially higher margins. Creative was the entry point and an added switching cost where it kept clients loyal and anchored the media AOR relationship that generated the real billings.

This chart, in other words, is showing you a business line that large agencies deliberately kept lower-margin to drive a higher-margin outcome elsewhere in the portfolio. It is not straightforwardly a story of value destruction. It is a story of a strategic trade-off, made consciously over decades, that happened to produce exactly the price-per-unit curve Farmer has spent his career documenting.

The problem is not that the trade-off was made. It is that it was made without maintaining the quality and strategic positioning of the creative work itself so that when media margins came under pressure from platform disintermediation, programmatic commoditisation, and AI-powered automated buying, agencies found themselves with a devalued creative capability and no premium pricing power anywhere in the portfolio. The creative line fell. The media line is now under threat. The strategy that required both is exposed.

Second step closer, and now we can see the mirage was never water at all. It was a portfolio line that the industry chose to let fall.

Here is the most important thing to understand about this chart: the trend will not reverse. And any strategy built on reversing it is a complete red herring. It's like complaining about gravity.

Take the final step. The mirage dissolves entirely.

Price per unit of creative work will continue to fall. AI will accelerate it. Production tools will get faster and cheaper. Creative volume will keep expanding. Every technology-driven industry follows this pattern. In semiconductors, the cost per transistor down more than 99.99 percent from 1990 to 2023 while total revenue grew roughly 10x. Computing. Telecoms. Digital storage. All the same. Romancing our industry as different because it is 'creativity' is itself a kind of mirage, the comforting belief that creative work is somehow exempt from the economics that govern everything else. It isn't. Price per individual unit falls continuously. Total value of the industry grows. Price per unit is not the lever and never was.

The question for every agency and marketing services business is not "how do we rebuild price per creative deliverable?" That question has no useful answer.

The question is: in the face of this gravity, which of three positions do you occupy?

1. Low-cost manufacture. You compete on volume and efficiency with AI-enabled production at scale, minimal overhead, commoditised delivery at competitive speed.

2. Quality differentiation. You compete on the nature of what is produced with creative work that shifts culture, builds genuine brand equity, and generates disproportionate commercial return. This commands a price premium.

3. Adjacent value capture. You move upstream into strategy, transformation, and the decisions that happen before any creative brief exists. This is where consulting firms have spent thirty years building. The pricing multiple tells the structural story: McKinsey charges five to six times the cost of its people; agencies charge roughly 2.2 times. That gap is not a negotiation failure. It is a positioning failure. The agencies that close the gap will not do so by arguing about deliverable pricing. They will do so by being in different rooms, advising on different decisions, long before the brief arrives.

The irony embedded in this chart is almost complete. Three decades of an industry devaluing human creativity, insight, and strategy in favour of the executional volume AI is now absorbing. The commodity layer is being automated. The question is whether the industry uses that liberation to finally charge properly for what it should always have been leading with.

The mirage is gone. The landscape underneath it is actually promising. But only for the agencies willing to see it clearly, rather than standing in the desert, thirsty, waiting for the water to come back.

Farmer's data is pointing at a real problem. But it's pointing at it sideways, through a metric that can only see cost-per-unit, from a dataset that only includes agencies with scope management problems, measured by a tool that its creator also sells the solution to. The chart is not wrong. It is being read in a way that leads to the wrong response.

The line will keep going down. Build your strategy around that fact, not against it. The agencies that thrive will be the ones who understand that the Farmer curve going to zero on commodity production is not a crisis. It is a door.

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 19 March 2026 in the Brandflow newsletter on LinkedIn.