A plan with one variable is an extrapolation

Most capacity plans model one number carefully and take the other one on trust.

There are two numbers in any plan for what your technology costs next year, and they are multiplied together. Usage gets a curve, a range, three scenarios, a week of somebody's attention. The price gets one figure, today's, repeated down twelve months with nothing beside it.

Nobody would call that a forecast if the subject were a car journey. Distance modelled properly, fuel at today's price. We would call it a distance estimate, and we would be right.

Why the second number goes quiet

It is not carelessness. A price you do not set reads like weather rather than like an input.

And compute got cheaper for long enough that "it falls, or at worst it holds" stopped being something anybody wrote down. It became the ground the plan stands on. Nobody argues with the ground. Call it furniture. Furniture is fine to build around, until the day somebody shifts it.

Twenty minutes, no context

Say a capacity plan lands in front of you and you have twenty minutes.

Do not open the growth curve. Somebody already spent weeks in there, and challenging it burns the meeting on the half that has had the most attention.

Go to the unit price and ask one thing: is this number allowed to move?

Both answers are useful. If it sits there once, at today's rate, with no range and no note, the plan extrapolates. It may be an excellent extrapolation. It is not a forecast. If the price already carries a range, somebody has been here before you, and the useful move is to find them and ask what changed their mind.

The difference is not accuracy

Nobody in the room can predict the price of compute. I cannot either, and that is not what separates the two documents.

A plan with one variable is an extrapolation. A plan with two is a forecast. Not because the second one is more accurate, and not because it uses a better model. Because it has written down what it is exposed to.

One row, and whose name is in it

Write the assumption down where the plan can see it. Underneath it, write what the total becomes if the price goes against you. One sentence, one row.

What makes that worth an afternoon rather than a minute is whose name ends up in it. The person who built the growth curve will fill it in as a footnote. The person who owns the budget will fill it in as a decision, because they are the one who has to live with the answer.

One more thing, because it surprises people. Writing the exposure down usually makes a plan calmer, not scarier. Often the move everyone is quietly worried about turns out to be one the estate can absorb, and you only find that out because somebody wrote the row.

The page below has the row on it. It carries blanks rather than examples, because the moment a figure sits in a template somebody inherits that figure, and nobody knows your estate better than you do.