One assay, day by day

This is a hypothetical walkthrough. It shows the shape of an Assay engagement; it does not describe an actual client or deal.

Two years ago you opened an office that never hit its numbers. The plan said it would break even sooner than it did, and the person who pushed for it left before the lease expired. Nobody brings it up in planning meetings anymore. They bring up discipline.

This year the question is a new office, this time because two of your largest customers are already building engineering teams in that market. When you raise it, the room agrees it is interesting and moves to the next item. Somewhere on the payroll is the person who knows whether it would work. In the meeting, that person watches your face.

So instead of asking the room again, you run an assay: a private, two week market on the decision itself, open to the fourteen employees you name.

You frame the decision as a single question with two paths: open the office, or don't. Both paths open at parity. The market starts with no opinion, and whatever it comes to believe, it has to be paid into believing: every trade is made with real money.

Each of the fourteen answers the question the room kept ducking, in the only currency that keeps people exact: their own expectation of what a share is worth on each path. A trade is an opinion with a price on it, and opinions can move; anyone can revise their position any day the market is open. None of it commits the company to anything. The prices advise, and nothing becomes actionable unless you decide to act.

For the first week, nothing moves. Prices drift a few cents apart, and the instrument reports exactly what it knows: nothing yet. That flat stretch is uncomfortable to watch, but it is also the point. A reading that cannot say nothing is not a reading worth trusting when it finally says something.

On day eight, someone begins selling the hold side and buying the act side, in a size that moves the reading. You cannot see who: while trading is open, no position carries a name, so conviction gets priced without anyone having to stand up and defend it. The reading moves to +14.6 percent in favor of acting. Two more accounts follow that afternoon.

By day twelve, the reading has climbed to +38.2 percent and holds there. Eleven of the fourteen employees in the market have taken a position. For two years the room treated a new office as a risk worth avoiding. Your own people, paid to be exact instead of agreeable, are now pricing caution itself as the more expensive path.

The market closes on day fourteen with that reading intact. An assay advises, it never binds, so you bring the number back to the same room that has been quiet on this question for two years and ask what, specifically, the skeptics know that the price doesn't. Nothing surfaces. You decide to act, and the office opens.

Set that next to a strategy engagement: three months and a fee that gets spent whether the study is right or not, against two weeks and a reading built from the people who already work there. One returns a deck. The other returns a price.