Evolution of agency-client relationships with AI

The agency world has 16 months left. Maybe less. WPP’s share price has halved since May - down 38% in six months. That’s not a blip, if you listen closely you'll hear the sound of an old model collapsing in real time. Why the collapse? Your clients are quietly bringing the work in-house, with the help of AI tooling. We’ve seen this with Unilever scaling its AI content studio across 18 markets (and adding 30% productivity gains in asset production) and TikTok launching Smart+ and Symphony to auto-generate entire ad campaigns. Agencies aren’t losing pitches, they've lost the right to be in the conversation in the first place. The reality today is: • A junior with toolset of Midjourney, Sora 2, Nano Banana and ChatGPT-5 and can out-produce £70K senior designers • Add Claude code, Cursor, Vercel + Supabase and you're out-producing a development team • AI-driven analytics and planning now outperform human insight teams on speed and accuracy. WPP’s board still talks recovery, but the market’s already pricing in the death of the traditional agency model. So what’s an agency to do? 1️⃣ Be brutally honest about what clients can do in-house. Stop selling production they can replicate with AI. 2️⃣ Double-down on genuine strategy. AI can’t replace business judgement or cultural intuition - yet. 3️⃣ Use AI in an advanced way. Design workflows, build agents, automate delivery. Don’t talk about AI - ship with it. We’ve been training teams in this for 18 months. The pattern we see it that agencies are still needed - just for different things. The old model is well and truly beyond resus: Creative concept → Design → Media plan → Traffic → Report The new model is thriving: Strategic insight → AI workflow design → Internal capability building → Outcomes Agencies can keep pretending it’s cyclical. Or they can start leading the revolution that’s already eating them alive. Follow me for more on navigating the collapse - and what replaces it. #AI #Agencies #work

We asked 400 agency clients what they want their agencies doing with AI. "Reduce fees for us" came 16th out of 20. AI-Opportunity Embracers are the largest group of agency leaders at 35%, and what defines them is a set of beliefs about their clients. 80% say clients expect their agency to lead the way in applying AI to marketing. 78% say clients expect help understanding and using AI effectively — against 52% and 47% across all agency leaders. That's a real bet on a particular picture of what clients want. We're in the unusual position of being able to check it, because we ask the clients too. The bet is a good one, and better than I expected. 94% of clients expect something to change at their agency in the next year because of AI, and what they're asking for is elevation: → Increase the effectiveness of our marketing — 52% → Improve the quality of strategy for us — 52% → Generate better new ideas — 52% → Develop more effective strategies for us — 49% → Create higher-quality work — 47% Teaching them to use AI is on that list too, at 37%. That's real demand and worth serving. It sits eleventh, below ten different versions of make-the-work-better. Then the bottom of the same list. Reduce fees for us, 16th. Cut administrative staff, 18th. Cut experienced strategists, 19th. Cut experienced creatives, 20th. Every option that saves the client money sits in the bottom quarter. And 82% of them say they will pay more for specialized expertise. Agencies have spent this year bracing for AI to become a price conversation. Their clients are having a quality conversation. So, Embracers: your instinct is right and your framing is costing you. Lead with what the client gets — sharper strategy, better ideas, higher-quality work.

The agency world has 18 months left. Maybe less. Sir Martin Sorrell's S4 Capital just reported an 11.4% drop in Q1 sales - a canary in the digital coal mine. It's not just economic headwinds. It's not just Trump's tariffs. It's that your clients are quietly bringing work in-house. With AI tools. We’ve seen this very publicly over the past few weeks with Dualingo, Shopify and Klarna all taking an AI first approach. The reality today is: • A junior with Midjourney can replace multiple £50-£70K designers • AI can draft campaigns faster than your team (who’s reinventing the wheel every time anyway) • AI analytics outperform your insights department Sorrell, ever the optimist at 80, promises things will turn around in H2. But the trend is clear: brands don't need agencies for tasks they can do themselves with AI. So what's an agency to do? 1. Be brutally honest about what clients can now do in-house Don't sell services they can easily replicate with AI. 2. Double down on strategic thinking AI can't replace genuine business understanding. 3. Use AI in an advanced way Create demand by using AI tools to deliver more value. We've been teaching teams how to use AI effectively for 18 months. The results? They still need agencies - just for different things. The old model is dying: Creative concept → Design → Media plan → Traffic → Report The new model is emerging: Strategic insight → AI workflow design → Internal capability building → Outcomes Agencies need to go from fighting this shift to leading it. Follow me for more on navigating the changes in the agency and job market created by AI.

People often ask if I miss the “good old days” of advertising. The short answer? No. The 15% commission model of the “old boys’ club” era was an oligopoly. Today’s market is faster, leaner, and far more demanding, but it is also vastly more authentic, transparent, and meritocratic. I recently sat down with Cecilia Garzella at Campaign US in our New York office to discuss the shifts reshaping our industry. As major tech firms dial back internal resources, we are seeing a rebalancing away from pure in-housing, with brands recognising how difficult it is to keep pace with technology across multiple verticals. At the same time, as platforms like Meta and Google automate ad buying, enterprise clients will never hand total control over to a single walled garden; the agency’s crucial role is evolving into that of an independent validator. Underpinning all of this is a fundamental restructuring of professional services. Based on our work with Professor Karim Lakhani at Harvard Business School, AI is driving a distinct two-tier model: fewer, highly compensated strategists and creatives at the top, supported by a massive, tech-enabled implementation engine below. As AI, quantum, and blockchain strip out historical supply-chain frictions, fee structures are moving permanently toward total transparency. Margins have tightened, and the macroeconomic climate demands relentless efficiency. But with global scale and constant technological evolution, the business is infinitely more interesting today than it was 40 years ago, if you are willing to adapt. Read the full interview in Campaign US here: https://lnkd.in/g5uMrDwB

Agency defensiveness is rarely about the tool. It is about what the tool means. I’m working with a large global CPG company on an AI marketing transformation. The moment new workflows started sitting outside the agency’s direct control, resistance showed up fast. Not resistance to AI itself. Resistance to what AI implies about how decisions get made, who owns what, and where value sits. That is the real disruption. Most brands are still layering AI on top of legacy agency models, then wondering why the impact feels incremental. The reason is simple: 📉 If you add AI at the end, you optimize the old model. 📈 If you build AI in from day one, you redesign the model itself. AI is not just an enabler. It is a design input for the future operating model. That also means brands need to be explicit about the role of the agency. The role is not to gatekeep tool selection. It is to operate brilliantly within the new structure and deliver outcomes. The best brands will not treat this as a fight between AI and agencies. They will use AI to redesign the model, then define clearly where agencies create the most value. Brands now have a real choice: 1️⃣ Bring agencies into the redesign early and reshape the model together, 2️⃣ Define the new model first and let partners adapt to it. Curious how others are thinking about this transition. #advertising #media #tech

The rise of AI doesn’t make agencies irrelevant. It makes them indispensable—if they evolve. I’ve heard the speculation: “With AI doing so much, will we need agencies anymore?” The truth is, AI isn’t eliminating the need for agencies—it’s exposing which ones are ready for the future. The partners embracing AI aren’t being replaced. They’re leading. They’re building smarter workflows, unlocking new insights, and evolving from executional support to strategic acceleration. After last week’s post on the 20-60-20 rule, many of you shared how you're balancing AI automation with human oversight. Today, let’s take that one step further—into how this balance is driving real operational excellence and unlocking new doors in advertising. Here’s a simple framework I’ve observed among the most successful AI adopters: Enhance → Automate → Innovate Enhance: Use AI to improve existing processes—make them faster, smarter, more scalable. Automate: Remove repetitive tasks. Free your teams to focus on strategy, not spreadsheets. Innovate: Use AI to unlock new capabilities that weren’t feasible before. Let’s bring this to life with a real-world partner example: A tech partner began by enhancing keyword research. What used to take hours, AI now does in minutes—suggesting keywords based on vast data signals. Next, they automated reporting. AI now builds reports with insights and recommendations pulled from multiple sources. Their analysts? They’re back to focusing on strategy. Then came innovation. By combining AI-driven audience insights with creative optimization, they built a system that dynamically adjusts ad content based on real-time performance. That level of personalization? Simply wasn’t possible before. Here’s the kicker: human expertise remained essential at every step. Keyword research still needed a strategist to align with brand goals. AI-generated reports required interpretation to guide decisions. And the personalization engine? It’s tuned and refined by creatives and planners every day. This brings me back to a core belief: AI is a collaborator—not a replacement. The partners winning in this space aren’t just using AI—they’re working with it to amplify their teams and build smarter solutions. Looking ahead? I see AI evolving from optimization to orchestration. Predicting trends. Adjusting strategies in real time. Maybe even composing full-funnel campaigns with inputs from multiple signals and channels. But we’re not there yet—and that’s exciting. Because it means there’s still time to build, test, and shape what this future looks like. So let me ask you: How has AI helped you enhance, automate, or innovate your operations? What new possibilities are you starting to explore? #AmazonAds #AI #FutureOfAdvertising

The agency black box has had a pretty good run. I think it’s time to open it. One of the things that attracted me to Horizon Media was Bill Koenigsberg’s commitment to transparency. If Horizon receives a discount or savings from a publisher, it goes back to the client. A simple idea, but an even more powerful standard. Now, I think we can take that philosophy much further. At Horizon, we’re working toward a model where clients can have a much clearer view of what’s happening with their marketing: how audiences are behaving, how campaigns are performing, and ultimately whether those media dollars are helping the business grow. AI makes it possible to connect systems and information that have historically been frustratingly disconnected. That transparency also changes the agency relationship. We’re shifting toward compensation tied to our clients achieving their business goals. That’s a harder model to build, but it gets much closer to the relationship I believe agencies should have with clients. I don’t want our ambition for AI to stop at doing yesterday’s work more cheaply. I want it to help us prove what works, make better decisions with clients, and put our success on the same scoreboard as theirs. That feels like a much more interesting future for the agency business.

Over the past year, I’ve been tracking a trend that’s now impossible to ignore. AI is not replacing comms jobs one-to-one. But it is replacing pieces of those jobs. And companies are reorganizing teams because of it. That’s the theme of my latest short video and the focus of the newest episode of 𝗧𝗵𝗲 𝗜𝗻𝘁𝗲𝗹𝗹𝗶𝗴𝗲𝗻𝘁 𝗖𝗼𝗺𝗺𝘂𝗻𝗶𝗰𝗮𝘁𝗼𝗿, which I’ve officially resumed. Here’s the data: 📉 𝟳𝟯 𝗽𝗲𝗿𝗰𝗲𝗻𝘁 of marketing leaders using AI have already reduced external content spend. 📉 𝟴𝟯 𝗽𝗲𝗿𝗰𝗲𝗻𝘁 say they expect to reduce agency spending further as AI tools improve. 📉 𝟭𝟭 𝗽𝗲𝗿𝗰𝗲𝗻𝘁 say they would eliminate agencies for basic content work once AI becomes “good enough.” And brands are saying the quiet part out loud. 🚫 𝗠𝗼𝗻𝗱𝗲𝗹𝗲𝘇 plans to cut agency fees by up to 50 percent. 🚫 𝗨𝗻𝗶𝗹𝗲𝘃𝗲𝗿’𝘀 𝗔𝗜 𝗦𝘁𝘂𝗱𝗶𝗼 is producing assets 30 percent faster. 🚫 𝗪𝗣𝗣 says brands using AI are reducing what they pay their agencies. And it’s showing up inside my own business. In 2024, about 80 percent of my clients were in-house comms teams. This year, 60 to 70 percent are agencies looking to adapt. 𝗧𝗵𝗮𝘁 𝗶𝘀 𝗻𝗼𝘁 𝗮 𝗰𝗼𝗶𝗻𝗰𝗶𝗱𝗲𝗻𝗰𝗲. Companies have spent the past two years pouring money into AI infrastructure and workflow redesign. One Nvidia chip costs as much as a car, and companies are buying them by the thousands. That money has to come from somewhere. So organizations are shifting dollars from lower-margin areas they believe have less strategic impact. And many have concluded that PR and comms fall into that category. 𝗔𝗴𝗲𝗻𝗰𝗶𝗲𝘀 𝗳𝗲𝗲𝗹 𝘁𝗵𝗶𝘀 𝗳𝗶𝗿𝘀𝘁. 🚫 Scopes are shrinking. 🚫 2026 decisions are being delayed. 🚫 Brands are internalizing work they used to outsource. This is not about AI replacing communicators. It is about the economics of AI reshaping the financial foundation of agency and in-house relationships. I’m looking forward to continuing this conversation on the podcast and here on LinkedIn, because this moment matters for every communicator, at every level. Would love your thoughts or feedback. How are you seeing this play out inside your organization? 👇🏾

What does the future of professional services look like when every client has AI at their fingertips? This isn't a hypothetical. Our data shows a majority of decision makers at large enterprises are deploying or getting ready to deploy agents within the next 12 months...for each major functional area of their business. Sure, these deployments may be small to start, but AI agents are building the war chests (and revenues) to scale them. $6B+ raised just this past year. Legal AI alone captured $1.25B. Coding agents hit $1.8B. Every dollar flowing into these tools is a direct challenge to the traditional advisory model. The existential questions professional services leaders (and enterprise clients) are asking: Why hire consultants when AI agents can deliver expertise on demand? And why hire consultants to build AI agents when the playbooks are still being written? Here's what our research uncovered: The billable-hour model is dying. Enterprise GenAI pilots are stalling at alarming rates. But the firms that will thrive won't be the ones clinging to yesterday's playbook. The future belongs to firms that evolve from advisors to builders and orchestrators. Drawing from CB Insights data on 100+ AI partnerships, investments, acquisitions and 60+ interviews with senior executives, we identified four moves that separate tomorrow's winners from today's casualties: ↳ Orchestrate the AI agent tech stack:  Don't just recommend tools. Build, integrate, and scale them. ↳ Activate proprietary data: Transform decades of client knowledge into fuel for smarter agents. ↳ Productize your services: Shift from custom engagements to scalable AI-powered platforms. ↳ Lead the human-AI workforce transformation: Redesign how humans and agents collaborate, internally and for clients. The winners aren't just advising on AI. They're building it, orchestrating it, and fundamentally reimagining their business models. The transformation is happening now with McKinsey already deploying 12,000 AI agents internally and Accenture repositioning their entire firm around AI building vs. advising. Our report, "The Future of Professional Services," is live and maps out in detail the 4 strategic pivots firms must make to thrive in the AI agent era. Read the full report here: https://lnkd.in/ezre-qWS

The Death of the Billable Hour: Why Pharma Agencies Must Reinvent Themselves or Risk Obsolescence The pharmaceutical agency landscape is undergoing its most dramatic transformation since the digital revolution. After decades of pharma and life sciences agency work, I've witnessed something unprecedented: AI isn't just changing how we work - it's making traditional life science agency models obsolete. The brutal reality? AI now executes 80% of core life sciences agency functions in 20% of the time. While agencies cling to billable-hour models, their clients are asking: "Why pay for 40 human hours when algorithms deliver superior results in 40 minutes?" Here's what's happening: 🤖 Specialist AI models produce regulatory-grade scientific content indistinguishable from human output 📊 Predictive analytics identify emerging KOL networks 6-8 months before human analysts ⚡ AI-powered compliance engines audit medical content with precision that reduces review cycles from weeks to less than an hour 🔍 Advanced systems cross-reference thousands of documents in real-time, ensuring narrative consistency The existential questions keeping life science agency leaders awake: ❓ What happens to agencies built on time-and-materials pricing when AI delivers the same output in minutes? ❓How do we justify our margins when clients can get 90% of our deliverables from AI at 10% of the cost? ❓What happens when clients start building internal AI capabilities and cut out agencies entirely? ❓How do we maintain revenue growth when our core profit centres are being commoditised? ❓Can we retrain our workforce fast enough before clients realise they don't need us for most tactical work? But here's the twist: This isn't about human vs. machine - it's about hybrid transformation. The agencies that will thrive are those reimagining themselves as AI-powered strategic orchestrators, where human expertise directs algorithmic precision to solve pharma's complex challenges. Three critical success factors: 1️⃣ Capability Liquidation & Reinvestment: Systematically replace legacy services with AI while redirecting resources to proprietary strategic offerings 2️⃣ Value Re-engineering: Shift from hourly billing to outcome-based models tied to commercial outcomes 3️⃣ Talent Evolution: Develop "AI Behaviour Architects" and "Evidence Orchestrators" - hybrid professionals who bridge AI capabilities with pharmaceutical domain mastery The transformation window is closing fast. First movers building proprietary AI infrastructure will establish unassailable market positions. Late adopters will find themselves trapped in a value erosion spiral. The bottom line: AI's greatest value isn't replacing human ingenuity - it's amplifying it at unprecedented scale. The future belongs to life science agencies that recognise this and act decisively. What are you seeing in your organisation? Do you need help transforming into a disruption-proof agency? #PharmaInnovation #AITransformation #DigitalHealth