Markets are not read. They are tested.

Every market analysis makes one promise: a picture.

And the picture usually arrives later than the decision. By the time the analysis is complete, either the opening has gone or the conditions it was written against have changed.

That is not a failing of the analyst. It is a failing of what we expect analysis to do.

A traditional market publishes nothing

In a traditional market, the information is not somewhere you can download. It is in invoices, in relationships, and in the memory of people who have spent years in that market.

The advantage is that whatever you get is real. Nobody built it to influence you.

The disadvantage is that it is slow, incomplete, and always arrives from someone with an interest in the matter.

An online market publishes too much

The online market has the opposite problem. There are plenty of numbers.

And an abundance of numbers creates a confidence that is not always earned. With a chart in front of you, it feels like knowing. What you may actually know is only what has been measured — not what matters.

A number that is correct but does not answer your question is more dangerous than no number at all, because it builds the decision on false confidence.

What they have in common

Neither the traditional market nor the online one tells you what happens if you enter.

Both describe the situation as it stands. Your decision is not about the situation as it stands; it is about something that does not exist yet.

No volume of analysis closes that gap.

What does close it

One small, real test says more than a complete report. Not because it is more precise, but because it is about the right question.

A good test has four properties.

It is small — small enough that failing it does not weigh down the next decision.

It is reversible. It does not cut the current route; it runs alongside it.

Its criteria are set in advance. Before you start, you know which number means continue and which means stop. Otherwise, once the result is in, a justification can be found for any result.

Its cost is accepted in advance. A test you have to fight for a budget for is not a test; it is a project.

And one thing usually left out

Market analysis without an analysis of the organisation's own decision path is incomplete.

You can read a market entirely correctly and still change nothing, because an analysis that meets the wrong decision path never reaches a decision.

So before asking what the market says, ask what route that answer is supposed to travel to become a decision. If there is no route, the answer is of no use.

In short

The traditional market says too little and the online market says too much. Neither tells you the future.

Rather than carrying analysis through to certainty — which it never reaches — carry it to the point of being enough to design a test, and ask the market itself for the rest.