Why is the first phase so small?

Occasionally a client is frustrated by this. They can see the whole thing, they have the budget, and they want to get on with it.

I still hold the line, and here is the reasoning.

Everything before software is a hypothesis

What everyone believes about how the process works is a theory until real people use real software on a real Tuesday. Small phases test that theory in weeks. Large builds test it at the end, which is the most expensive possible moment to find out you were wrong.

You do not know if I am any good yet

References and case studies help, but nothing tells you what a supplier is like to work with except working with them. A small first phase is the cheapest way to find that out.

If I am slow, or vague, or I do not listen, you learn it for a modest amount rather than four months into something large.

What small must not mean

It must not mean useless on its own. If phase one is only valuable once you buy phases two and three, that is not a small start, it is a deposit dressed up.

Every phase should stand alone. If you stopped after the first one, you should still be better off than before.

The exception

If there is a hard external deadline — a trade show, a regulatory date, a contract start — then we plan differently and I say so. But the default is small, live, and useful, then next.