Transcription
By 2028, we'll double profits and have the people. That's a global holding company CEO talking to Forester on the record about what's coming for hundreds of thousands of agency employees. 15% of agency jobs eliminated this year. Not over the next decade, this year. 47,000 positions gone in 12 months. And I'm going to show you who's on the kill list.
Forester just released their 2026 predictions. This comes after agencies already cut 8% of jobs in 2025. The bleeding is accelerating, not slowing down.
I'm Jeff Sour. Along with my partners, I built an agency, grew it 1,000% in 5 years, made the Inc. 5000 list five times, and sold it for 8 figures. So, when Forester says 15% of agency jobs vanish, I don't need to guess what that looks like. I watched it happen during the digital disruption from 2008 to 2012. And I've spent months compiling over 2,000 pages of research on what's coming this time. And I can tell you this is worse. It's faster. And most people in the industry have no idea which side of this they're standing on. If you work at an agency, freelance for agencies, or you're watching this industry from the outside trying to figure out if there's opportunity here, subscribe. I'm going to be covering the shift all year because it's moving faster than anybody expected.
Let me tell you exactly what Forester said because the original prediction wasn't scary enough. Then they made it scarier. Back in 2022, Forester projected that 7.5% of US agency jobs would be automated by 2030. About 32,000 positions over 8 years. That felt manageable, gradual, something the industry could absorb. Then in late 2025, they threw that out. New prediction, 15% eliminated in 2026 alone. They didn't just move up the timeline, they doubled the severity and compressed eight years into one.
Jay Padisol, he's the VP and principal analyst at Forester, described what's happening as a complete transformation. His words, agencies are no longer acting as agents. They're becoming marketing purveyors selling products, technology, and media rather than service hours. And that sounds abstract until you understand what it actually means. The traditional agency model is built on one thing, selling time. You hire people, you build their hours, you mark up the labor. That's how agencies have worked for a century. Now, AI breaks that model. When AI can do in seconds what used to take a junior team member a week, you don't need a junior team. And when you don't need a junior team, you don't need the middle manager supervising them. And when you don't need either, you start asking why you're paying for the office space they sat in. The holding company CEO who said double profits have the people. They're not being cruel. They're just doing the math.
Now, here's what's wild about this moment. Agency CEOs aren't hiding it anymore. They're saying it publicly. Arthur Sadun, CEO of Publus, the best performing holding company right now, told investors that 80% of revenue from their media business now comes from AI-driven work. 80%. He said, "AI handles everything from how clients plan and buy media to how campaigns are optimized and measured." His exact words, "The tech is doing the heavy lifting." Poopus grew 5.6% last year while other holding companies declined. They beat their peers by about 700 basis points and they did it without adding headcount. In fact, they still cut around 200 people. Even the winner is cutting.
Over at WPP, headcount dropped from 110,000 to 104,000 in about a year. 7,000 people gone. At the same time, 85% of their client-facing staff are now using their AI platform. Notice the correlation. Headcount down 7,000, AI usage at 85%. The new WPP CEO called 2025 performance unacceptable and brought in McKenzie to advise on the transformation. Revenue is expected to decline another 5 to 6% this year.
Omnicom just merged with IPG and they're targeting 750 million in cost savings. The CFO is pretty clear about where that comes from. Labor costs always lead the balance sheet. They went from 128,000 combined staff to about 105,000 in roughly one year. and they just retired three iconic agency brands. DDB, FCB, and Mullen Low. DDB was founded in 1949. FCB was founded in 1873. These brands survived world wars. They survived the Great Depression. They survived the dot crash and the 2008 financial crisis. These brands didn't survive this change.
So, who's actually getting cut? Forester broke it down. 28% of job losses are clerical and administrative. data entry, scheduling coordinators, expense reporting, admin assistants, the stuff AI handles almost trivially now. 22% are sales and business development roles. They're junior people who research prospects, drafted proposals, coordinated outreach. AI does that faster and doesn't need a salary. 18% are market research positions. Survey analysis, data synthesis, report writing. An agency that needed 10 researchers can now deliver comparable output with three using AI tools.
But those are just the percentages. Let me get specific about roles. Junior copywriters, AI generates thousands of headline variations in seconds. The apprenticeship model where you learn by writing low-level copy, that's broken. There's no on-ramp anymore. Media planners, Sir Martin Sorell, who founded WPP and now runs 54 Capital said something that should terrify anybody in media. there are 250,000 people in media planning and buying. There won't be 250,000 jobs in 2 to 3 years. His argument is that media buying becomes completely algorithmic. Performance Max, Advantage Plus, the platforms do the targeting, allocation, and bidding automatically. What about production artists and junior designers? AI image generation lets you resize assets, remove backgrounds, create variations instantly. Tasks that required teams of juniors are now one-click operations. Account coordinators, scheduling, routing, status tracking. AI handles all of it.
Forester calls this the workforce inversion. Historically, agencies made money through labor arbitrage. You had expensive creative directors overseeing cheap junior talent. You had expensive account directors managing through junior coordinators. The juniors did the volume work, the seniors supervised, and the spread was profit. AI inverts that completely. You don't need the junior layer anymore. The model flips to small teams of senior talent working directly with AI assistants. Forester's exact language, the more creative and original the agency role, the less likely it'll be replaced. Originality is the most significant factor that lowers a job's automation potential. The entry-level path into the industry is disappearing, and the mid-level roles that supervise entry-level work are disappearing with them.
Now, I'm curious. If you're watching this and you work in the agency world, drop a comment and tell me which side of this you think you're on. Are you in a role that's clearly vulnerable, or do you feel relatively protected? I read every comment and I'm genuinely trying to understand how people in the industry are experiencing this.
Here's the part that doesn't get talked about enough. It's not just that big agencies are slow to adopt AI. It's that their entire business model makes adaptation almost impossible. Take a 1,000 person agency. You're carrying somewhere between 50 and 75 million in annual overhead. Real estate, management layers, HR, legal, finance, technology, infrastructure. That's 15 to 20% of gross revenue before you've spent a dollar on actual project work or taken any profit. So, you win a $500,000 engagement. Direct costs are 300,000 and that leaves $200,000 in contribution. Minus your allocated overhead, call it 100,000, and your actual profit is maybe 100,000. That's 20% on a good day. Now, revenue dips 10%. Suddenly, you're losing money on every project because overhead doesn't flex. You can't just turn off the lease or fire your CFO because business slowed down.
Small operators don't have this problem. A micro agency with one principle, a couple of contractors, and AI tools has overhead in the tens of thousands. They can profitably serve clients at rates where large agencies would lose money. Then there's decision speed. A solo operator or a threeperson team can test a new channel, adopt a new AI tool, pivot their entire strategy in hours. A large agency, that same decision requires alignment across creative, media, account management and strategy. Then you need security validation from it. Cost modeling from business ops, legal review of the vendor contract, compliance signoff from corporate, days become weeks, weeks become months. By the time you've implemented, the tool has evolved or competitors already integrated it.
WPP announced a major NVIDIA partnership, invested heavily in AI. Their market share still declined. Pulicus achieved margin expansion through AI, but didn't gain share either. The investment isn't translating into competitive advantage because by the time big agencies deploy, the capability is table stakes.
But here's the part that really kills them. Even if they solve the AI problem internally, clients are still walking out the door. The Association of National Advertisers reports that 82% of their members now have in-house agencies. That's up from 58% in 2013. And 65% have moved established business from external agencies to in-house in just the past 3 years. Gartner's CMO survey found that 39% of CMOs plan to cut agency spend this year and 22% explicitly said AI reduced their need for external agencies. They're not just cutting budgets, they're saying out loud that AI lets them do the work themselves. Deote surveyed retail executives and found that 94% expect to bring more marketing inhouse. 75% are reducing reliance on external agencies. This is a deliberate strategic choice not cyclical belt tightening.
So big agencies are getting squeezed from both directions simultaneously. Inside AI is eliminating the roles that they use to profit from. Outside clients are pulling work in-house because AI makes it possible for brands to handle capabilities they used to need agencies for. There's the Kodak analogy here that's almost too perfect. Kodak actually invented the digital camera. They literally created the technology, but they couldn't pivot because their business model depended on selling film consumables with recurring revenue. Agencies have invented AI capabilities. Publoist invested 12 billion euros. WPP has all these platforms, but their business model depends on selling hours. Headcount is how they make money. Of course, we know Kodak filed for bankruptcy in 2012. The same year, Instagram sold for a billion dollars with 13 employees.
Agencies that survive this won't be slightly smaller versions of what exists today. They'll be fundamentally different operators. Which brings us to the part that actually matters. Who survives this? And is there opportunity here if you're positioned for it? The economic comparison is stark. Traditional agencies operate at 15 to 20% profit margins. AI enabled micro agencies, and I'm talking about real operations, not theoretical ones, are running at 50 to 80% margins. Revenue per employee at a traditional agency 150,000 to 200,000. At an AI native small shop, 500,000 to over a million. Annual overhead at a 50 person agency is 300,000 to 500,000 just to keep the lights on. A micro agency with its AI stack, $3,000 to $12,000. These aren't incremental differences. This is an order of magnitude.
Let me give you some concrete examples. DesignJoy is a oneperson subscription design agency run by a guy named Brett Williams. He makes roughly $2 million a year with zero employees. Productized model, AI assisted production, no overhead. Myar won Ade small agency of the year in 2025. Their revenue grew 50% year-over-year. They doubled their team while retaining 91% of staff and 95% of clients. They added Netflix and BJ's Restaurants as clients, a small and focused agency competing for and winning major brand accounts. There's a New Orleans shop called DAA Media that jumped from number 4538 to number 2150 on the Inc. 5000 in a single year. Triple digit revenue over three years under 150 employees. And on the bigger independent side, Wromote just acquired Giant Spoon. Combined, they have 700 employees and manage over three billion in media spend. Clients include HBO, Google, TransUnion. They won ad ages performance marketing agency of the year. That's the PE backed independent model as an alternative to holding companies.
Small operators have five structural advantages that big agencies simply cannot replicate. First is overhead. A micro agency can profitably serve clients at price points where large agencies would lose money. The math doesn't work for them. It only works for you. Second is decision velocity. You can test, adopt, and pivot in hours. They take months. Third is client attention. Small operators provide direct access to senior talent. Large agencies assign junior staff to standard engagements and reserve senior people for premium accounts. Fourth is technology adoption. You can integrate new AI tools immediately. They have legacy systems, governance requirements, and deployment timelines measured in quarters. Fifth, and this one matters more than people realize, no fake work. Big agencies often create work to justify retainers, meetings that don't need to happen, decks that don't need to exist. That's how you fill hours when you're billing hours. Small operators sell the result. When your client is paying for the outcome, you're aligned with them instead of aligned against them.
Look, I want to be clear about something. The industry isn't dying. The traditional agency job is, though. Ad spend is growing 9.5% this year. The IAB just confirmed that the money is still there. In fact, there's more money than ever, but it's flowing to fewer, more efficient operators. The question isn't whether marketing services have a future. The question is whether you're positioned to capture the new flow or standing in the path of the old one.
If you're currently an agency watching this happen, you've got a decision to make. Do you wait and hope that your role survives or do you start rebuilding the skills and positioning that puts you on the winning side? If you're outside the agency world and you see opportunity here, you're not wrong. This is the biggest opening in a generation for people who can deliver marketing outcomes without the traditional overhead structure.
I've built something called the 99 services list. It's a breakdown of specific productizable services that small operators can deliver profitably using AI. The services that clients actually need and are willing to pay for right now. If you want access to that list, drop a comment below with the word services and I'll send it to you. And if you want to go deeper on the shift, how to position yourself, what skills matter, how to actually build something on the winning side of this, subscribe and hit the bell. I'm covering this all year because it's moving so fast and most people are not paying attention. That's it for this one and I'll see you in the next.