Write the strategy so it can be proved wrong.
A pitch is built to survive questions. A thesis is built to be killed by one — which is the only kind worth writing.
SEAM's thesis names what would have to be true in risk order with the cheapest test for each, alongside a deliberately adversarial bear case and a graveyard analysis of eleven dead companies. It produced one falsifiable threshold that now governs the model: revenue per account per year, where nothing below $50 survived and nothing above $150 failed.
Most strategy documents are written to be persuasive. They assemble supporting evidence, handle objections, and arrive where the author already was. Such a document cannot be wrong, which means it cannot teach you anything — its only function is to move a decision that had already been made.
The alternative is to write it so a single finding can kill it. That means naming what would have to be true for this to work, ordering those assumptions by how badly you lose if they are false, and specifying the cheapest test for each. It means writing the bear case as well as you can rather than as weakly as you can get away with. And it means studying companies that died doing something adjacent, without reaching for the reasons you are different.
What comes out the other side is usually not a conclusion but a threshold — a number that decides things later, when you are tired and motivated to see what you want to see. For SEAM that number is revenue per account per year: below fifty dollars nothing in the graveyard survived, above one hundred and fifty nothing failed. That single band now rules out whole product shapes without another meeting.
The discipline that makes this work is marking every claim as fact with a source or as inference, and treating “could not verify” as a finding rather than a gap to fill later. An unverified assumption you have labelled is an open risk. An unverified assumption you have smoothed over is a landmine with your own handwriting on it.