Method
The method, not the résumé
Every write-up on this site follows the same shape, on purpose. Here it is, named plainly — not a case-study slideshow with logos on it, because there isn't one to show honestly. Just the five things this practice actually does to a decision before calling it made.
The five steps
-
Name the fork before you argue for a side
Most decisions get presented after the fact as if there was only ever one sane option. There wasn't. Before defending a choice, the actual alternative that was live at the time gets written down — not a straw man built to lose. If a decision can't survive being paired with its real alternative, it wasn't a decision, it was a preference wearing a decision's clothes.
The real alternative Not a straw man -
State the tradeoff in one sentence, before committing
If a decision can't be finished as "we're giving up ___ to get ___," it isn't understood yet — it's just been picked as a favorite. Every proof point on this site states its tradeoff in exactly one sentence for that reason, not as a caption written after the fact but as the test the decision had to pass first.
See it in the build-vs-buy write-up → -
Write down what would flip the answer
A decision that can't be wrong under any circumstances isn't a decision, it's dogma. Each write-up on this site ends by naming the condition under which the opposite call would have been correct. Naming that condition is what proves the reasoning was actually load-bearing, rather than a conclusion in search of a justification.
See it in the ad-targeting write-up → -
When the stakes are real, argue with yourself a second time
For anything with real exposure — money, a filing position, a system other people depend on — a single pass isn't enough. The second pass's job isn't to re-check the arithmetic, it's to try to disprove the first pass's own conclusions. Most of what it catches isn't a new defect, it's the first pass's own overconfidence.
See it in the audit write-up →Argue to disprove it Re-check the math -
Publish the miss, not just the win
A track record that only shows the parts that went well costs more credibility than it earns, because there's no way to tell it apart from one that's simply never been checked. The corrected-cost-model proof point on this site is on the site because it was wrong, not despite it.
See it in the proof section →
This is the same method behind every proof point on the site — not a separate practice kept for show. If you want it run against a decision you're actually facing, email directly.