Major Decisions Deserve an Objective Diagnosis First
The evidence that reaches a major decision was assembled by people who hold a position on it. An objective diagnosis is the description of the business written before anyone needed it to say something.
A major decision is usually made against the company's description of itself.
Somewhere behind every major decision sits a document that describes the company. A board pack, a strategy deck, a forecast with a narrative page. The decision itself, a market entry or a reorganisation or a platform rebuild, is argued against that description. Nobody in the room asks when the description was last checked against the business.
The two drift apart, and the drift is structural. Descriptions are written for audiences, and audiences reward coherence, so the account acquires a clean storyline the operation does not have. Products that were quietly deprioritised stay in the deck. A customer segment that stopped renewing survives in the segmentation. The company on paper is the company as it intended to be, several intentions ago. Nothing about that is negligent. Keeping the account current is nobody's job, and the account is only tested when something large leans on it.
The pattern in the field
A payments business had decided, in principle, to rebuild its checkout platform. The case rested on a description of checkout as the group's growth engine, and every number in the paper was accurate. The operational record told it differently: growth had moved to a newer product line more than a year earlier, and the platform earning the rebuild was in managed decline everywhere except the document. The decision was not wrong because the analysis was sloppy; it was wrong because the thing being analysed no longer existed. Once the description was corrected, the rebuild re-scoped itself in a week. The correction was the work.
By the time a decision is major, every input has an author with a position.
The reflex explanation is politics, and it is mostly unfair. People do not lie to decision papers. They advocate, which is their job, and advocacy shapes evidence long before anyone decides anything. The decision paper is the end of a long supply chain, and every station on it has preferences. By the time material reaches the room it has usually been shaped three times over.
- Produced by the sponsor. The analysis behind an option is commissioned by the person proposing it, and scoped to the question that favours it.
- Filtered on the way up. Each summary layer removes a caveat, because caveats read as weakness and summaries are written to survive meetings.
- Timed for the room. Data that lands the week of the decision arrives too late to be checked, and borrows credibility from the deadline.
None of this is misconduct. It is what advocacy does to information, and it happens in companies with honest cultures and excellent people. Objectivity does not fail at the moment of decision; it fails months earlier, in the ordinary process by which evidence is assembled. The room can be scrupulously fair and still be choosing between two well-argued fictions.
When facts pick sides
Once options are on the table the effect accelerates. Every fact in circulation gets enlisted by one side or the other, and a fact with a side is no longer available as a check. This is the same mechanism that lets a consulting brief harden around its least evidenced sentence: the description stops being a hypothesis the moment someone needs it.
An objective diagnosis is a description written before anyone needs it to say something.
The correction is a sequencing rule, not a virtue. Before the options are framed, the company writes down what is true: what is happening, since when, what is funded as opposed to endorsed, what customers do as opposed to say, which numbers moved and which were reclassified. The description is agreed as accurate by the people who will later disagree about what to do — and it is agreed first, while it is still nobody's ammunition.
What the description draws on
The sources are records nobody wrote for the argument, which is exactly what qualifies them. The calendar and the budget, read the way the work before the work describes. The renewal ledger rather than the pipeline narrative. Support queues, cohort behaviour, the last three attempts at the problem and the assumption each one carried. Interviews belong in the file too, read as a record of what people are willing to say rather than of what the organisation does.
The test of the finished document is symmetry. A description is objective when it would read the same whichever way the decision later went. If a paragraph only makes sense as support for one option, it is advocacy that arrived early, and it goes back.
A description written to support a decision cannot be used to check it.
Who can hold the pen
Someone without a position, which inside the company is rarer than it sounds. A major decision reaches into most functions, and whoever is senior enough to write the description is usually senior enough to hold a stake in it. The pen can go to an internal function with nothing at risk, or outside altogether. What matters is that the author's next year does not change with the outcome. An owner can hold it; a sponsor cannot.
Run in reverse, the same exercise produces ammunition.
The common failure is not the absence of diagnosis. It is diagnosis commissioned after the decision, to dress it. The reorganisation is announced, then a review is run to establish why it was necessary. The acquisition is emotionally closed, then diligence is scoped to confirm the price. The resulting document looks identical to an objective diagnosis and does the opposite work. It is the corporate equivalent of writing the conclusion first, and it is more common than either of the honest failures.
The tells
They are structural rather than tonal. The deadline exists before the question does. The scope excludes the option that lost. The author reports to the sponsor. Findings circulate to be agreed rather than argued with. Any one of these means the description sits downstream of the decision, and a description downstream of the decision protects nothing.
The cost arrives later and off the books. The decision proceeds with its risk unexamined, which is survivable; companies absorb bad decisions. What compounds is what the exercise teaches: that description is a formality performed after choosing. The next honest diagnosis, including one that would have stopped a loss, reads as obstruction, because the organisation has learned what these documents are for.
Reading the business before framing the decision is where the thinking becomes work.
None of this argues for slower decisions. A settled description speeds the argument up, because the room stops litigating facts and starts weighing trade-offs, which is the part that needs the room. Deciding remains the leadership's job; the diagnosis only cleans the ground it happens on. The wider family of these patterns sits under diagnosis before solution.
Where we do this alongside a leadership team, the reading runs a few weeks ahead of the decision being framed. The output is short: a description of the business the decision must survive, agreed as true by people who want different things from it. Sometimes it changes the decision, occasionally it dissolves it, and either way the options that reach the table are options about the company that exists. The framing and the choosing stay with the people who own them; the consulting work is the reading, not the verdict.