Most bad decisions are not bad judgment. They are a mismatch between the size of the choice and the size of the process. Decision-making frameworks exist to fix that mismatch: small, reversible choices get a two-minute rule, and large, hard-to-undo choices get a written process.
The practical move is to sort first and decide second. Before you think about the options, decide which kind of decision you are holding. That single step eliminates most of the time people lose to deliberating over things that do not deserve deliberation.
Sort the decision before you make it
Two questions do almost all the sorting work.
Can I undo this cheaply? If reversing the choice costs you an afternoon and no money, it is reversible. If it costs a lease, a relationship, or a year, it is not.
Does the downside have a floor? A choice where the worst case is "I wasted £40 and a Saturday" has a floor. A choice where the worst case keeps getting worse does not.
Reversible with a floor: decide in minutes, alone, with no spreadsheet. Irreversible without a clear floor: use a written process and take days. The error almost everybody makes is applying the slow process to the first category and the fast process to the second.
Jeff Bezos popularised the related idea of one-way and two-way doors in his shareholder letters, and the framing holds up well outside business: walk through two-way doors quickly, and slow down at one-way doors.
Framework one: the two-minute default
For reversible, floored decisions, set a standing rule and stop reconsidering it. Which restaurant. Which of two similar books. Whether to take the earlier train.
Pick a default and apply it mechanically:
- When the options are close, take the one that is easier to exit.
- When you genuinely cannot tell, take the cheaper one.
- When you have deliberated for two minutes, take whichever you were leaning toward at the thirty-second mark.
This is not carelessness. Close options are close by definition — the expected difference in outcome is small, so the time spent choosing is the dominant cost. Spending twenty minutes to improve a coin-flip is a net loss even when you pick correctly.
Framework two: the weighted criteria grid
For medium decisions with three or more real options — which course to take, which city to visit for a month, which of two job offers — a weighted grid converts a fog of preferences into a number you can argue with.
- List your options as columns.
- List four to six criteria as rows. No more than six, or the weights stop meaning anything.
- Give each criterion a weight from 1 to 5 based on how much you actually care, not how much you think you should.
- Score each option 1 to 5 against each criterion.
- Multiply, sum, and look at the result.
| Criterion | Weight | Option A | Option B |
|---|---|---|---|
| Cost | 4 | 4 (16) | 2 (8) |
| Time commitment | 5 | 2 (10) | 4 (20) |
| Fits existing skills | 3 | 5 (15) | 3 (9) |
| Easy to exit | 4 | 3 (12) | 5 (20) |
| Total | 53 | 57 |
The number is not the decision. The value of the grid is that it exposes the weights, and weights are where people quietly lie to themselves. If the total picks an option your gut rejects, do not override it silently — go back and find the criterion you weighted dishonestly. That is usually the real finding.
Framework three: the pre-mortem
Before committing to an irreversible choice, write the failure story. Gary Klein’s pre-mortem technique asks you to assume the decision has already gone badly, then explain why.
Set a timer for ten minutes and write, in past tense: "It is six months from now and this clearly did not work. Here is what happened." Be specific. Name the thing that broke first.
Then read it back and look for failures you can cheaply prevent. Most pre-mortems surface two or three risks that take an hour to mitigate and would have taken a year to recover from. The point is not to talk yourself out of the decision — it is to make the version you commit to a better one.
Framework four: the regret horizon
For decisions that are mostly about identity and direction rather than logistics, the comparison grid tends to produce confident nonsense. These are the choices where the criteria themselves are what you are unsure about.
Ask the question at three horizons:
- Ten days. What will I think of this when the disruption is at its worst?
- Ten months. What will I think once the novelty and the inconvenience have both worn off?
- Ten years. What will I wish I had done, looking back?
Suzy Welch’s 10/10/10 framing is useful because the three answers usually disagree, and the disagreement is informative. If ten days says no and ten years says yes, you are looking at a short-term cost problem, not a direction problem — solve the cost rather than abandoning the choice.
Matching the framework to the decision
| Decision type | Example | Framework | Time to spend |
|---|---|---|---|
| Reversible, low stakes | Which of two similar tools | Two-minute default | Under 2 minutes |
| Several real options, comparable criteria | Which course to enroll in | Weighted criteria grid | 30–60 minutes |
| Irreversible, committed resources | Signing a 12-month lease | Pre-mortem | A few days |
| Direction and identity | Changing field entirely | Regret horizon | Weeks, with notes |
Keep this table somewhere you will see it. The sorting step is the part people skip, and it is the part that saves the most time.
Five failure modes to watch for
Gathering information you will not use. Before researching, write down what answer would change your choice. If no answer would change it, you have already decided and you are stalling.
Treating a reversible choice as permanent. Most commitments have an exit. Finding the exit before you commit often converts a hard decision into an easy one.
Adding options to avoid choosing. A fifth option rarely helps and reliably resets your deliberation. Cap the list at three or four and close it.
Asking people who will not bear the cost. Advice is cheap in proportion to the adviser’s exposure. Weight input by who absorbs the downside.
Deciding when depleted. For anything above the two-minute tier, make the call in the first half of your day. Tired decisions default to whatever requires no change, which is a choice disguised as a non-choice.
Build a short decision log
Keep one page with four columns: the decision, the date, the framework you used, and what you expected to happen. Add a fifth column for the actual outcome and fill it in later.
Within a few months the log tells you something no framework can: where your judgment is reliably good and where it is reliably off. Some people consistently overestimate how much time they will have. Others consistently underestimate how much they will dislike an obligation. You cannot correct a bias you have not measured, and a log is the cheapest measurement available.
Review the log quarterly. You are looking for the gap between expected and actual, not for whether the decision was "right" — a good decision with a bad outcome is still a good decision, and a lucky one is not a method.
Start with one tier
Do not install all four frameworks this week. Pick the tier where you lose the most time. For most people that is tier one: a long tail of small, reversible choices soaking up attention that should be going to the two or three decisions that actually matter this year.
Set the two-minute default, hold to it for a fortnight, and notice what you do with the recovered attention.
For a broader toolkit you can pull from as different decisions come up, start with The Top 50 Self-Help Frameworks page — several of the entries there are decision tools you can run on a single sheet of paper.