Why we don't do predictions
We publish no forecasts, no probabilities and no accuracy rates. This is a design decision, not a hedge.
1. Predictions are unverifiable in the only timeframe that matters
A client makes one decision, once. There is no control group, no counterfactual, and no way to separate a good outcome from a good decision. Selling certainty into that situation means being judged on luck.
2. Every serious adviser works from structure, not certainty
Lawyers advise on enforceability. Accountants advise on treatment. Investment advisers advise within a mandate. None of them promise an outcome — and the categories that do are not comparable engagements.
3. The useful layer is verification
When a decision goes wrong, the cause is usually visible beforehand: an assumption nobody checked, a control nobody defined, a payment that moved before diligence finished. These are checkable facts, not prophecies — and they are what a review sets out in writing.
4. It changes the client relationship for the better
Clients who buy a framework check the reasoning. Clients who buy certainty demand outcomes. The first relationship is one we can sustain; the second is one nobody can.
What we deliver instead
- What the decision depends on, in writing
- Which assumptions are verified and which are not
- Two or three paths, with the trade-offs between them
- The irreversible points, and what to settle before you reach them
If that is what you need, start with the twelve questions or read what a structured decision review is.
What a review can be held accountable for
A review can be checked: was each assumption identified, was each verification listed, was the control question answered, were the irreversible points named? That is a verifiable deliverable, and it is the standard we accept. A prediction cannot be checked in the same way, because the client makes one decision, once, with no counterfactual.
How this shows up in the deliverable
| Assumptions log | Every assumption stated, with what would verify it |
| Control map | Who can block, when, and how deadlock resolves |
| Sequencing | What must be true before each commitment of capital |
| Exit conditions | Notice, valuation and restrictions if the relationship deteriorates |
If that is the standard you want from an adviser, the checklist is where to start.
What we publish instead of a forecast
- Which assumptions carry the decision — logged individually, so each one can be checked.
- What would have to be evidenced — the verification list that belongs in your negotiations.
- Who can block the decision, and when — the control question, answered from the documents.
- Where the irreversible points sit — the steps after which reversing costs a renegotiation.
How a review is produced: four steps
- Step 1 — Scoping. A short conversation to establish the decision and its boundaries; a scoping request is answered within 24 hours.
- Step 2 — Intake. The decision, the parties, the capital, the constraints and the documents.
- Step 3 — Analysis. Four dimensions, two or three scenario paths, and the verification list.
- Step 4 — Delivery. A written memo over WhatsApp in 48 hours, with one follow-up exchange answered within 72 hours of delivery.
Numbers that hold whatever happens next
These are the only figures we publish, and none of them depends on an outcome: a scoping request is answered within 24 hours, a memo is delivered in 48 hours, and one written follow-up exchange is answered within 72 hours of delivery. Everything else in a review is either a fact to verify or a mechanism to define in the agreement.
Why the question still comes up
Clients ask for a probability because it would simplify the decision: one number, one threshold, decided. The request is reasonable and the answer is still no — a probability attached to a single, non-repeatable decision cannot be checked afterwards, and an unverifiable claim is worth less than a list of things to verify. What can be checked is whether the assumptions were identified and whether the irreversible points were named.
What this costs you
Nothing above replaces a lawyer's reading of the documents, and we say so at scoping. The value is in the ordering: what to verify first, what to turn into a condition, and which questions to send to counsel already framed. If that ordering is not useful for your decision, the twelve questions will tell you in twenty minutes.
