The demo is fun. The second use is the interesting one.

A quick demo is a lovely thing. An idea that existed only in my head starts doing something on a screen. AI makes that moment easier to reach, and I’m enjoying the possibilities.

To understand whether the idea is useful, I want to get to the next question: would someone come back and use it again?

Start with a real piece of work

Imagine a tool that prepares a weekly report. The first demonstration uses tidy sample data. The next use has a renamed column, a missing week, and someone who needs the output in a slightly different format.

Those details tell me whether I’ve understood the job. A clear message about missing data may matter more than another chart.

Give the product a narrow promise

I want to be able to explain what the tool does in one sentence. Then I want its first use to deliver that result without a tour of everything it might become.

AI can help me explore layouts and simplify the words. I need to check that a person can find the next step, understand the result, and recover when something goes wrong.

Watch for a reason to return

For an early prototype, I’d ask someone to try it on a real task using appropriate, permitted data. Where do they hesitate? What do they still do by hand? Do they have a reason to use it next week?

There’s room to learn that the idea needs changing. Building quickly makes another attempt more affordable, provided I listen to what the first attempt showed me.

I’m interested in small products that earn their place. If one saves someone time or makes a recurring job easier, it gives me a clue about which possibility to explore next. That’s worth finding out.

— Vince