The Friday afternoon problem

A multi-million turnover business asked me to build them a custom AI system from scratch. The real fix took one week.

They wanted bespoke enterprise software, built around their exact workflows. It sounded like a serious project. Custom integrations, months of development. But before I quoted anything, I asked one question: what is the task that eats the most time every single week?

Every Friday, someone on the team spent half the day copying invoice numbers from one system into another and sending follow-up emails. Half a day, every week, doing the same thing in the same order.

The solution was a workflow fix. A simple AI brief and a saved template handled the copying and the follow-ups. One week from scoping to running. The custom build they had imagined was never needed.

Schoolyard rules apply. Pick fights you know you can win. Your first AI project should be a small, low-risk task that repeats every week. One win teaches you more than any platform. You see how AI works in your business and where it falls over.

I keep seeing this play out. The business that starts with one boring, repetitive task and nails it is the one that knows exactly what to automate next. The one that starts with the big vision is still scoping months later.

Once you have one win, you know exactly where to aim the second one. Start with the problem that shows up every Friday.

Comment STARTER and I'll send you the 5B Starter Kit. The full framework, the prompts, and a 7-day calendar to get your first AI win.