Inside a strategy session.
A paid working session in which your deployment gets modelled properly before any capital moves, including the scenarios where it does not work. You leave with the model and its assumptions written down.
3 min readWhat it is.
It is a scheduled hour on your actual numbers, using the planning tool I built for this. Your capital, your entity, your timeline, the hardware that is realistically available, and the facility capacity that actually exists.
It is not a sales call with a spreadsheet open. The output is a document you keep, and a fair share of these end with me saying the deployment does not make sense at the size or the timing you had in mind. That is a legitimate outcome and it is worth paying for. Finding it out afterwards costs considerably more.
The fee and what it includes are on pricing.
What gets modelled.
The fleet
Machine model, quantity and efficiency in joules per terahash, against what can actually be bought and delivered rather than a specification sheet.
The power rate
Modelled at $0.08 per kilowatt hour. That is the market rate for hosted capacity, checked against competitors’ published figures, and it is deliberately not a promotional number.
Difficulty growth
Modelled at 10% a year and stated as an assumption. Fitting a trend to recent history returns figures near 37%, which is plainly wrong: that period included difficulty falling, and compounded over fifteen years it implies a rise of more than a hundredfold.
The horizon
Five years. Machines reach end of life around sixty months, and running a longer projection mostly adds years in which the fleet is dead and the chart still climbs.
Tax treatment
Entity type, the depreciation provisions available, and the limits that decide whether a deduction lands in the year you want it. The conclusions belong to your accountant.
Operating costs
How hosting gets paid, and whether that means selling production on a schedule. This changes the holding period on part of the output and it needs deciding rather than discovering.
The assumptions are stated, and they are meant to be argued with.
Every input above is written into the document. If you disagree with one, change it and see what happens, which is the point of having a model rather than a conclusion.
I am specific about this because the category norm is the opposite. A model tuned to a favourable power rate, an optimistic difficulty path and a fifteen year horizon will produce a very attractive chart, and none of the three will be visible on the page it is printed on. A projection is only as good as the assumptions it is hiding.
The corollary is that my own numbers are open to the same test. Ask what happens at a lower price, a higher difficulty growth rate, or a shorter machine life. If the answer is bad, that is information you wanted.
What it does not produce.
Worth being explicit, because these are the things this kind of document usually leads with.
No return projection
I do not publish or promise a rate of return, a payback period presented as fact, or a production figure you can rely on. The model shows scenarios, including ones where the deployment loses money.
No dollar tax figure
I will not put a number on your tax outcome. It depends on your bracket, your entity, your state and your involvement, and stating it would mean announcing your accountant’s conclusion on their behalf.
No uptime promise
Hashpower coverage, where it is taken, is a remedy for downtime rather than a promised level of production.
No recommendation to proceed
The session is not structured to end in a yes. Ending in a clear no is a successful outcome and I would rather reach it in an hour than in eighteen months.
What you leave with.
A written model with the inputs on the face of it, the scenarios run out including the unfavourable ones, and the Bitcoin price the plan needs in order to work. That last number is the most useful thing in the document and it is the one most operators leave out.
It is written to be handed to your accountant and read by someone who was not on the call. If you decide to go ahead, it becomes the reference the deployment is executed against. If you decide not to, you have a documented reason, which is worth more than a vague sense that it seemed complicated.
Model it before you commit anything.
An hour on your capital, your entity and your timeline, with the assumptions written down and open to challenge. If the deployment does not make sense at the size you have in mind, that is what the session will tell you. Start with a discovery call if you would rather talk first.
Book a strategy session Paid session. Assumptions stated. A genuine answer, including no.