The biggest AI mistake? Starting with the solution.

One thing has really stood out to me over the past few weeks. After speaking with companies across insurance, travel, energy and customer service, I've noticed the same pattern again and again. Companies start by evaluating AI vendors, demos and use cases, and only afterwards ask themselves where AI could actually create meaningful business value.

To me, that's the wrong order.

Instead of first agreeing on the business challenges that matter most, the discussion quickly becomes: "Which of these solutions could we use?" The result is often a portfolio of interesting AI initiatives that don't build on each other, don't follow a common vision and make it difficult to prioritise investments.

I've become convinced that AI strategies should be built the other way around. Start with the business. Understand the problems worth solving. Prioritise them. Build a roadmap. Only then decide which technologies and partners help you get there.

Otherwise, you're not choosing the best solution.

You're choosing the best demo.

Technology should support the strategy - not define it.

LINK