Most AI initiatives don’t fail because the technology isn’t ready. They fail much earlier, at the point where teams choose what to automate first.
Leaders are usually advised to “start big” or “pick something transformational.” In practice, that leads to workflows that are too broad, too cross-functional, or too politically loaded to move quickly. The pilot drags on, momentum fades, and AI quietly gets labelled as “not quite there yet.”
The reality is simpler. The first AI project should not be the most ambitious. It should be the most contained, measurable, and outcome-driven.
In our latest episode of The AI Opportunity and the accompanying blog, we explain the method we use with real businesses to choose first workflows that actually make it to production. It focuses on identifying small, high-friction processes where a single team can see clear impact fast, creating proof of value before scaling.
If you’re deciding where to start with AI, or trying to rescue an early pilot that’s stalled, this approach will save you a lot of unnecessary trial and error.
Short on time? Listen to the key insights on Spotify | Apple Podcasts | YouTube or read the blog