One of the hardest parts of advising clients on AI isn’t the technology. It’s choosing the right first project.
Most AI initiatives fail before they start, not because AI cannot deliver, but because the initial workflow is too big or too vague. As a domain expert, clients often look to you and that decision often sits with you. It matters more than any model or tool choice.
From the work we’ve shared in The AI Opportunity podcast and blog, there are three criteria that consistently predict whether a client’s first AI project will succeed.
-
Impact: The workflow should matter. High dependency processes where improvement is felt immediately across revenue, cost, or risk are ideal for this.
-
Clarity within 30 days: If a project cannot be clearly scoped, implemented, and evaluated inside a month, it usually turns into a time sink. Tight timelines force focus and prevent AI projects from drifting into chaos.
-
Measurable: Clear metrics to measure the before and after effects. These don’t need to be perfect, but it does need to be credible and clear to the clients. Without this, even a successful project struggles to build belief.
If you are helping clients think through their first AI use case and want to pressure test whether a workflow is the right place to start, feel free to reply in this thread or message us directly.
Or if you had a similar experience with your clients successful or otherwise we would love to hear about it!