The Free Half

Ten years ago, the phone was the free part.

You signed a two-year contract and the handset cost you nothing. The carrier gave it away to win you, and made its money somewhere you barely thought about: the texts and the minutes. A text cost an operator a fraction of a penny to carry and sold for pennies each. Voice was the same. The thing you actually wanted, the phone in your hand all day, was the loss leader. The margin was in the messages flowing across the network for the next twenty-four months.

Then a free app arrived. WhatsApp reached 400 million users on the strength of 35 engineers and carried messages for nothing. iMessage did the same inside the phone itself. And the thing the operators had built their entire economics around went to zero.

Here is what happened next, and it is the whole point of this piece. Once the messaging margin collapsed, the carriers could no longer afford to bury the cost of the handset inside it. The free phone stopped being free. So they unbundled it and made the device its own line, financed openly over two or three years and moved their margin to data, to bundles, to enterprise. The free part and the paid part swapped places. The phone became the thing people covet, queue for, upgrade in cycles and finance like a car. The texts became the thing nobody pays for at all. The carriers saw it coming and rebuilt the model around it. They disrupted themselves before they were disrupted.

Most of the advertising industry is standing in exactly the place the operators stood in 2013. And it has not done the same thing.

I have spent more than a year writing about a middle that is dying, specifically the holding-company agencies caught between the integrated giants above them and the specialists below. Roughly 25,000 jobs came out of that middle in a single year. Omnicom, Dentsu, pre-merger IPG, Havas. The headlines have a ready explanation for all of it, and the explanation is wrong. Victims of AI.

AI is the perfect culprit. Invisible, ethereal, everywhere and nowhere, capable of being blamed for anything because it can be pictured doing everything. It is the scapegoat the whole industry has quietly agreed to use, because the alternative explanation is far more uncomfortable.

AI did not do this. AI is an accelerant thrown onto a fire the industry lit itself, years before the first usable model shipped. The flaw is structural, it is self-inflicted, and it is old. The holding companies built their scale in the wrong place. The wrong half.

And the failure is no longer theirs alone. It continues to reset the price of the work for everyone.

Those in the deathly middle built their infrastructure and their margin around two activities: production and media buying. Those were the engines. That is where the scale lived and where the money was made.

And to win those two engines and keep them, they gave away everything else. Growth strategy. Human insight. The creative idea itself. Priced at close to nothing (and often genuinely nothing) handed over to anchor the production contract and secure the media account.

It worked because the margin lived in the mystery, not in the value. You could charge handsomely for production while making things was a dark, capital-intensive craft clients could not do themselves. You could charge a media commission while buying media was a closed world of rate cards and relationships. The fee was never really attached to what was worth the most. It was attached to what clients could not yet see how to do without you.

And to be fair, it was not stupidity. For thirty years it was a brilliant, rational model that printed money. The flaw stayed invisible precisely because it was so profitable. Nobody dismantles the thing paying for the building.

This is the cruelty of it.

The work they handed over for nothing is precisely the work that does not automate. The strategy. The insight. The creativity. The short version: AI matches patterns, and creativity breaks them; it has no taste, no sense of which idea is worth making before the data exists; and it has no convictions, because conviction means holding a position that costs you something, and a model has nothing to lose.

Now hold that against the model.

Every loss-leader arrangement in commercial history gives away the cheap thing to sell the valuable one. Razor handles to sell blades. The holding companies did it backwards. They put the one durable, irreplaceable, genuinely human thing in the loss-leader slot and built their margin on the part a machine was always, eventually, going to be able to do.

So they are now losing on both sides at once. They cannot charge properly for the strategy and the creativity, because they trained the market for thirty years to expect those for free. And they can no longer make their margin on production and media, because those are being automated and in-housed out from under them. The half they gave away, and the half they sold are both gone at once.

There are exactly two ways out of the deathly middle, and the bifurcation everyone is watching is simply the sorting of companies into one or the other.

Above the middle: the integrated giants. Publicis. Accenture. The comfortable story is that they cracked the problem and finally worked out how to value creativity and strategy. They did not. They still under-charge for the human half exactly as everyone else does. What they did was move upstream and find new things to meter: a technology stack priced as software, first-party media priced as access, tooling priced on subscription. They did not learn to charge for the half they give away. They found new things to charge for instead. And it is worth saying plainly, because this is not advocacy for them: there is no guarantee the new revenue lasts either. The agentic shift could commoditise it the way the last engine was commoditised. Nobody has solved the original problem. They have outrun it.

Below the middle: the specialists. The boutiques and independents who win on depth, not breadth. They escaped by a different route, and it is the more honest one. They became superlative at one thing. A craft, a discipline, a single act done better than anyone and when you are recognised as the best in the world at something, you can finally charge for that thing. They never relied on scale where the value wasn't created. They priced the value directly. That is the one model in this entire industry that actually values the human half correctly, and it is available only to those willing to be genuinely, narrowly excellent.

The middle can do neither. It is too big to be a specialist and too undifferentiated, and too under-invested, to build what the giants built upstream. Caught between a capability ceiling and a distinction floor.

I am sometimes told I am too hard on the legacy underperforming holdcos, and by the time you read this there may be some comments below to that effect. I genuinely wish I weren't. There are friends and former colleagues across all of them, and there is no pleasure in any of it.

But it has to be said, because it is true. These businesses carry a terminal flaw in the way they are built. They have not invested enough to get upstream and price new things, and they are not good enough at any single thing to charge for it the way a specialist can. So as automation keeps eating into the activities they still depend on, they have no exit from the spiral. Not because they are badly run by bad people. Because the structure has a poison in it that predates the current management, and predates AI.

And here is the part that should worry everyone, not only them.

A business in that spiral, desperate for revenue and out of good options, will do the one thing that buys it time: it will take work at a low margin. Then at no margin. Then, to keep the lights on and the headcount busy, at a negative one.

When the largest players in a market continue to accept work below the cost of doing it well, they reset the price of everything around them. They train clients, all over again, that the work is worth even less than they were already paying. The desperate middle is not just dying. On the way down it is dragging the floor of the entire industry's pricing with it.

The floor of what creative work is worth cannot rise while there is desperate capacity in the market willing to give that work away. The clearing is brutal. It is also the mechanism. The price of the creative industry can only recover once the businesses that have spent a decade deflating it stop being able to.

And on the other side of that clearing is something genuinely hopeful, not as a platitude, but as the direct economic consequence of everything above. If the valuable half is the surviving half, then automating the other half is a gift to anyone with the nerve to take it. The Great Compression is not the death of the creative economy. It is the bill finally arriving for thirty years of mispricing and the chance to write the next invoice correctly.

This is what "more human than ever" actually means. Not a comforting slogan. A pricing strategy. The only defensible margin left in this industry is the human one: the taste, the conviction, the idea worth having. And for the first time in a generation, the economics are pointing back towards it.

So, briefly, for anyone leading through this:

1. Stop defending the texts… whatever your texts are.

The activity you make your margin on today is the activity most exposed to automation. Protecting it is buying time, not building a future.

2. Itemise the value.

The strategy and the creativity that have always travelled free on the invoice are the things to price first, not last.

3. Choose your exit.

Upstream, to new things you can commercialise or downstream, to a craft you can be the best in the world at. The one option that no longer exists is the middle.

The carriers survived because they saw which half of their business was about to be worth nothing, and rebuilt around the other half before they were forced to. They let the texts go and made the handset the thing worth paying for. They disrupted their own model before the model disrupted them.

Next week, tens of thousands of marketers and agency peeps will walk the Croisette to celebrate the work, the ideas, the craft and the convictions that their own industry's commercial model spent thirty years giving away for free. And somewhere along that same stretch of sand, on a yacht, one of them will hand exactly that to a client to win the business, and quietly pull the price of it down for everyone else.

The work on those stages is the Free Half. It was always the valuable half. The only question left is whether the industry can find the nerve to start charging for it, before the desperate few drag the floor out from under all of us.

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 15 June 2026 in the Brandflow newsletter on LinkedIn.