Peter sits down with Ammann Badlani — a 22-year veteran of digital media and commerce, currently building agentic systems inside the world’s largest agency networks — to talk about why the agency model is shifting from creative service to software, why platform black boxes keep failing CMOs, and why the bottleneck for performance teams is no longer execution capacity but orchestration logic.
Ammann’s core argument: the AI edge in agency work isn’t a single clever agent. It’s a team of narrow agents with a brain on top — and a willingness to rewrite the client’s brief before you ever hit run. The agencies that still treat AI as a faster version of the old workflow are the ones losing pitches to those rebuilding the workflow from scratch.
About Ammann Badlani
Ammann Badlani has spent 22 years in digital media and commerce — across paid search, SEO, programmatic, retail media, and shopper marketing. He started on the client side at Philips Electronics, ran his own consultancy for a decade, serving SMBs, and then moved into the agency world 13 years ago. He has led practices at Omnicom, GroupM, and WPP Media, working with brands including Ford, Unilever, Bose, and Church & Dwight. He launched Meridian Commerce Partners, a venture focused on AI-native commerce orchestration. He is based in Brooklyn, New York.
You can find Ammann on LinkedIn or email him at amman (at) meridiancommercepartners (dot) com
.
About Peter Benei
Peter Benei co-founded AI-Ready CMO the daily intelligence platform for senior marketing leaders. Peter has been serving as a CMO, marketing leader, and consultant to high-growth B2B scaleups for the past 10+ years. He has a background in advertising, working with Fortune 500 brands.
Connect with Peter on LinkedIn or read his newsletter.
Top 10 Takeaways
Adopting AI isn’t enough — the systems around it are the moat. “I’m using AI” was a credible answer in 2023. It isn’t anymore. The differentiator is the framework: how you structure agents, how data flows between them, and how outputs ladder back to a business question. Tools are a commodity. Workflow architecture is not.
The agency model is shifting from creative service to software. WPP and the rest aren’t pivoting because they want to — they’re pivoting because clients are demanding outcomes the old retainer-and-headcount model can’t deliver. The next decade of agency work looks more like productized orchestration than billable hours on a brief.
Black-box platform dashboards fail because clients still want to see inside. Google and Meta keep pitching the “trust the algorithm” model — give us the goal, we’ll handle the rest. Clients don’t buy it. They want transparency on where media shows up, what’s working, and what’s not. Any agentic system you build has to surface reasoning, not hide it.
No single platform sees the full picture, which is why orchestration is the opportunity. A CPG brand runs across TV, programmatic, retail media, search, shopper, influencer, and else. None of those silos sees the others. The agency that builds an agentic layer on top of all of them — pulling category sales, retailer performance, weather triggers, competitive moves into a single brain — does work that no platform can replicate.
The 20-agent team beats the one-agent generalist. Most marketers building with AI create a single agent that does too much. The right architecture is the inverse: many narrow specialists (category sales, retailer performance, search velocity, competitive moves) that talk to each other, with an orchestration brain on top translating the conversation into the language of a senior client.
Don’t push the brief down to every agent. Push it to the brain. If you narrow data collection at the agent level based on a half-baked client brief, you’ll miss the surprises that make the work valuable. Feed agents wide data; let the orchestration layer apply the brief. That’s how you find the insight the client didn’t know to ask for.
Briefs are starting points, not contracts. Clients often don’t know what they want — they know what they think they want. Your job isn’t to ignore the brief; it’s to rewrite it more precisely than they did, then go find the data that answers the question they actually have. The hand-holding is the value.
AI gives mid-market teams access to talent they couldn’t afford before. Big agencies used to staff Oxford-trained statisticians on the floor to find anomalies in performance data. Mid-market couldn’t. Now, an analyst persona inside an agentic system does the same pattern recognition at a fraction of the cost — and runs continuously, not in monthly cycles.
Hourly billing breaks the moment AI compresses the work. Updating 150 banner variants used to take a designer two hours. Now it takes minutes. If you’re billing the same way, you’re either lying or you’re losing the contract. The value shifts from execution time to hypothesis quality — what to test, what to read into the results, what to recommend next.
Build your own frameworks. Don’t just read other people’s. Ammann reads eMarketer, Euromonitor, and YouGov — but treats them as inputs, not outputs. The actual practice is synthesizing what he reads into three or four personal frameworks he keeps refining as new information lands. The frameworks are the asset. The articles are the raw material.
5 Things Worth a CMO’s Attention
1. The agency model is bifurcating, and it’s happening faster than the org charts suggest
The WPP earnings story isn’t an isolated event — it’s the leading edge of a structural shift. The big networks are being asked to deliver software-grade orchestration on creative-service contracts, and the math doesn’t work. Smaller, more nimble agencies that build proprietary agentic stacks are starting to win pitches against firms ten times their size, because the work product is genuinely different.
