Five steps, then a decision that is yours.

What you make, roughly how many people sell, and what system holds your customer records. A few minutes, and it is not a discovery call.

It exists for one reason: so that whatever gets measured later can be compared against something instead of floating on its own. It is the same set of questions every time, for every company, which is what makes the comparison mean anything.

One session, on a video call. There is a fixed set of commercial situations — a quote that went quiet, a deal that stopped moving, a buyer who changed jobs, an account that stopped ordering, a change at a customer that nobody saw in time. I ask which of them occur here, how often in your own words, and when the most recent one was.

I do not tell you which is worst. Your ranking is the finding, and supplying it would destroy the thing I came to measure. If nobody can answer for one of them, that is a finding too — it means the situation cannot currently be seen.

You pick a real one, with a date and names. You share your screen and your people drive. We walk it from the first moment somebody could have known, to whatever finally happened — every system it touched, every person it passed through, every point where a fact had to be typed twice or asked for again.

Some of what I count cannot be got from a conversation, which is why this is a screenshare and not an interview. I am counting what is there, not asking anyone to remember.

Not in the room, and not on the call. I read what you actually have against what a correct version of that work looks like, and the gap between the two is the finding.

This is deliberately not done live. A number produced under time pressure in front of the person who owns the process is a number nobody trusts afterward, including me.

What broke, where, and what that one instance cost. Every figure with the method that produced it, so you can check it rather than take my word.

What could not be captured, and why. If your records cannot answer a question, that is a finding rather than a blank — and it is often the most useful thing in the report.

And what this cannot tell you, said before you ask.

If you stop after step five you still have the trace, the finding and the number, and none of it required you to buy or change anything.

If you want the fix put in place, that is the next piece of work. I show your people what good looks like, we set it up in the systems you already own, and then I measure before and after against definitions that do not move. You run it afterward, because you are the one who has to live with it.

A single trace shows you where that instance broke. It does not prove the pattern, and I will not present it as though it does.

Whether it is the pattern is a larger piece of work across several situations, and I will tell you plainly which of the two you are looking at.

Nothing here has been run with a client yet. When it has, what I learned will change this page.