Is This Actually a Business Decision? A Test Worth Applying First
Some meetings circle the same ground for an hour and end exactly where they started. Everyone contributes, nobody is unreasonable, and no conclusion arrives.
The usual diagnosis is poor chairing or too many people in the room. Occasionally that is true. More often the problem is that the thing on the table was never a decision in the first place.
There is a test for this, and it takes about ten seconds.
The Test
A decision requires actionable alternatives. Two or more courses of action you could genuinely choose between, each leading somewhere different.
Compare two questions. "Should we build a new website or overhaul the existing one?" is a decision, because there are two paths and each has consequences you can investigate. "Which content management system is best?" is not, because best is undefined and nobody has named the alternatives.
The second question feels like a decision. It has a question mark, it concerns something important, and people will hold views about it. But there is nothing to choose between until someone converts it into options.
That conversion is the skill. Guidance on what to outsource demonstrates it well, replacing a vague sense that something should come off your plate with a specific question about what you would do with the recovered hours.
What Usually Fails the Test
Three types of question turn up in decision meetings without belonging there.
Requests for information disguised as decisions. "How much would a rebrand cost?" is research. It may inform a decision later, but it cannot be decided in a room.
Questions with only one real option. If the alternatives are proceeding or continuing exactly as you are, and continuing is untenable, you are planning rather than deciding.
Questions about preference rather than action. "What do we think of the new positioning?" invites opinion without requiring anything to follow from it. Useful conversation, not a decision.
None of these are wasted questions. They just belong in a different meeting, and treating them as decisions is what produces the circular hour.
Converting a Question Into a Decision
The fix is usually to name the alternatives explicitly, which is harder than it sounds and is most of the work.
Take "we need to free up my time", a sentiment most founders will recognise. That is not a decision, and it becomes one only when it turns into a choice between named options with different consequences.
Once the alternatives exist, you have something you can weigh. Before they exist, you have a mood.
What to Do Once It Passes
A question with real alternatives can be mapped, and mapping it is what stops the discussion looping.
The standard tool is a decision tree diagram, which uses three components. A decision node, drawn as a square, marks a point where you choose, while a chance node, drawn as a circle, marks an uncertain outcome you do not control.
An endpoint marks where a path finishes and nothing further needs deciding.
You work left to right from the root decision, branching at each choice and each uncertainty, and continue until every line reaches an endpoint with nothing further to decide. Rejected options stay visible on the map rather than disappearing, which matters when someone asks in six months why you did not do the other thing.
The value is not the drawing. It is that a mapped decision exposes which branches you have actually thought about and which you have been quietly assuming away.
Apply the Same Questions to Every Option
This is where most comparisons fall apart, and it is easy to miss in the moment.
If you ask whether one option would be easy to implement, you have to ask it of the others too. Otherwise you are not comparing options, you are building a case for the one you already prefer while subjecting the alternatives to scrutiny you never applied to it.
Consistent criteria are also what make a decision repeatable. A business that faces the same kind of choice every quarter, whether about hiring, pricing or which project to fund next, benefits enormously from evaluating each instance against the same questions rather than starting fresh and arguing from instinct.
Putting Numbers On It, Carefully
Where you have figures, a mapped decision supports arithmetic.
Assign a probability to each uncertain outcome, with the probabilities from any single point totalling 100 per cent. Assign a value to each endpoint. Then work from right to left, multiplying each outcome's value by its likelihood and subtracting the cost of the path that reaches it.
Where you do not have reliable probabilities, do not invent them. Rating outcomes as high, medium or low likelihood is honest and still useful, whereas a precise-looking number pulled from nowhere makes a guess look like analysis.
Risk appetite matters too. Two people reading identical numbers may reasonably choose differently, one preferring a modest certain return and the other accepting volatility for a larger upside. The map informs that judgment rather than replacing it.
Know the Limits
Mapped decisions have two failure modes worth naming.
They grow. A straightforward question can balloon into something unreadable, at which point a narrower view focusing only on the critical decisions and inputs serves better than an exhaustive one.
And errors cascade. A wrong assumption early in the map contaminates every branch downstream of it, which is why the assumptions at the root deserve more scrutiny than the detail at the edges.
Most importantly, a decision map runs on expectations. It cannot account for what nobody anticipated, and it is only ever as good as the information behind it. Anyone who has built a plan that a genuinely unforeseen event demolished already knows this.
That applies to bigger strategic calls as much as small ones. Building an AI strategy, for instance, involves weighing initiatives whose payoff is genuinely uncertain, and a map helps precisely because it forces you to state what you are assuming.
Conclusion
Before the next decision meeting, apply the test. Are there two or more things we could actually do, and does each lead somewhere different?
If yes, map it, apply the same questions to every option, and put numbers where you honestly have them.
If no, you have a research task or a conversation, and naming it as such will save the hour that would otherwise go in circles.
FAQ
1. What makes something a decision rather than a question?
A decision requires at least two actionable alternatives leading to different outcomes. A question with no named options, or only one viable path, is research or planning rather than a decision.
2. When is a decision tree worth drawing?
When a choice has multiple stages, uncertain outcomes you do not control, or consequences worth weighing systematically. It is particularly useful for decisions you face repeatedly, where consistent criteria reduce the influence of instinct.
3. What if I cannot estimate probabilities?
Use relative ratings such as high, medium and low rather than inventing figures. A map with honest uncertainty is more useful than one with precise numbers nobody can defend.
4. Can a mapped decision become too complicated?
Yes, and it is a common failure. When a map grows unreadable, narrow it to the critical decisions and inputs rather than expanding it further, and check the assumptions at the start before trusting the detail at the end.
