The half hour that decides everything

The owner has given you thirty minutes between two things that matter more. What you do with it determines whether there is a second conversation.

The thing that kills it is talking about what is now possible.

Nobody wants to hear what AI can do

They have been told. By a vendor, by their accountant, by a nephew. It has been vague every time and nothing came of it, so the topic now carries a small amount of accumulated scepticism you did not earn.

More importantly it is the wrong direction. A conversation that starts with capability makes them evaluate your technology. A conversation that starts with their week makes them describe their business, which is the only thing you actually need.

Ask what took up their week, and where somebody is staying late. Then stop talking.

The questions that work

Five, in roughly this order, and none of them mention software:

What does a normal week look like here? — orientation, and it is easy to answer.
What took up more time than it should have last week? — specific and recent beats general and abstract.
Who stays late, and what are they doing? — finds the irritation, and gives you a name.
How many of those do you do in a week? — the volume question, asked once you know what "those" are.
What happens when one is unusual? — the exception rate, which is your entire delivery risk.

Then the only question that matters at the end: may I sit with them for a day and watch?

Have the conversation with three of your five

Three, because the first one will go badly and you should not spend your best prospect on it.

Go in person, at a quiet hour — not Monday morning, not the end of the month. Say what you are doing plainly: learning, building something real, not selling anything today. That framing is generous to them and it happens to be true, which is why it works.

They start describing what they want built

Ten minutes in, the owner says: 'What we really need is an app where the drivers can log their hours.'

Conversations that produced a shadow day

Of the businesses you spoke to, how many agreed to let you sit and watch. This is the only outcome that matters from this stage.

What you heard

Three owners have described their week to you. That is more market research than most people do before building.