WPP's Chief AI Officer: AI-Built Marketing Solutions Can't Differentiate Your Business

Daniel Hulme, Chief AI Officer at WPP, has noticed a spreading phenomenon: employees using AI to quickly build HR systems, CRMs, and even marketing platforms—what's called vibe coding. His judgment is blunt: these solutions will never make your business different. It's not that the technology fails, but that they aren't solving the problems that set you apart.

Hulme draws an analogy from quantitative finance to explain why these self-built solutions are doomed to mediocrity. If you write an algorithm that can predict stock price movements based on historical data, and then start trading on it, here's the problem: the new behavior you've created didn't exist in the past data, so you can't predict the future based on the past. Marketing works the same way. If you can predict human behavior and then deploy ads that change that behavior, your entire infrastructure is out of date. This isn't an engineering problem—it's a structural self-referential bind.

His recommendation is to focus AI on the problems that truly differentiate your business and leave generic functions to specialist third parties. The complexity of the marketing supply chain demands the right data, the right talent, and algorithms that adapt rapidly to a changing world—competing with organizations that specialize in that area is very difficult. Vibe coding is good for proof-of-concepts, but anyone who has built software can see the unhappy paths: security, error handling, scaling. That's also why most projects never go beyond a minimum viable product.