The Decision-Making Process: 7 Steps That Actually Work
The decision making process is the sequence of steps a person or group follows to make a choice: defining the decision, finding and comparing options, checking the favorite for errors, committing, and reviewing the result. A good process scales its effort to the stakes and ends with a review, so each decision improves the next one.
- Category
- Decision process
- Origin
- Classic rational models in management theory; this version builds on the Mentarium Decision Loop
- Best used for
- Important choices with several options and real uncertainty
- Related
- Decision matrix, pre-mortem, decision journal
What is the decision-making process?
The decision-making process is the set of steps you follow to get from “I need to choose” to a committed choice, and then to a lesson from how it turned out. Most textbook versions have five to eight steps. They differ in names, but nearly all include defining the problem, finding options, evaluating them, choosing and following up.
The version below has seven steps. It is the Mentarium Decision Loop (Frame → Map → Stress-test → Decide → Review) with two practical steps added: sizing the decision before you start, and gathering only the information that could change your choice. Those two steps are where most people either waste effort or skip it.
A quick overview:
| Step | Question | Output |
|---|---|---|
| 1. Frame | What am I actually deciding, by when? | One sentence, a deadline, a definition of success |
| 2. Size | How much process does this deserve? | A Decision Weight Score |
| 3. Map | What are all the real options and their costs? | At least three options |
| 4. Gather | What would change my mind? | Answers to two or three key questions |
| 5. Stress-test | How could my favorite go wrong? | A pre-mortem and a Bias Radar check |
| 6. Decide | What do I choose and what do I expect? | A written prediction and confidence % |
| 7. Review | What happened vs what I predicted? | A lesson for the next decision |
Step 1: Frame the decision
Framing means stating exactly what is being decided, by when, and what a good outcome looks like. It sounds obvious. It is also the step most likely to go wrong, because a vague or narrow frame leads every later step in the wrong direction.
Write three lines:
- The decision, as a question. “Should we hire a second engineer this quarter?” not “the team situation.”
- The deadline. When does this need to be decided, and what happens if you don’t decide?
- What success looks like. In a year, how would you know this went well?
Then check the frame for two common traps:
- The yes/no trap. “Should I take this job?” hides other options: negotiate, take it later, ask for a different role. A whether-or-not question usually means you haven’t framed widely enough.
- The wrong-problem trap. Ask “why does this decision exist?” The question “should we redesign the website?” might really be “why are fewer visitors signing up?”, which has many more possible answers. The five whys is a fast way to find the real question.
Step 2: Size the decision
Not every decision deserves seven steps. Before you invest effort, score the decision with the Decision Weight Score (DWS), a 1 to 10 measure of how much process it needs.
Rate each factor from 1 to 3:
- Reversibility: 1 easy to undo, 2 costly to undo, 3 permanent.
- Impact: 1 small, 2 meaningful, 3 life- or business-changing.
- Uncertainty: 1 mostly known, 2 some unknowns, 3 mostly unknown.
DWS = round(R × I × U ÷ 27 × 10), minimum 1.
- 1 to 3: decide fast, under 10 minutes. Skip to step 6.
- 4 to 6: a light pass of every step, about an hour.
- 7 to 10: the full process, a pre-mortem, a written journal entry, and a night’s sleep.
Say you are deciding whether to move your savings into a different index fund. Reversibility 1 (you can move it back), impact 2, uncertainty 2: 4 ÷ 27 × 10 rounds to 1. Decide fast. Now say you are deciding whether to buy a house: reversibility 3, impact 3, uncertainty 2. That is 18 ÷ 27 × 10, which rounds to 7. Full process.
The Decision Weight Score calculator does the math. The value of this step is permission: permission to decide small things quickly, and a nudge to slow down on big ones.
Step 3: Map the options
Mapping means listing your options, including the ones not on the table, and what each one costs. Aim for at least three. Research on organizational decisions by Paul Nutt found that decisions framed as a single “yes or no” option failed more often than those where several alternatives were considered, which matches common sense: if you only consider one option, you are not really choosing.
Ways to find more options:
- Combine: can you do part of A and part of B?
- Sequence: can you do A now and B later?
- Test: can you try a small, reversible version first?
- Remove the option you prefer: if it vanished, what would you do? That answer is often a real contender.
For each option, write its cost, and use opportunity cost: the real price of an option includes the best alternative you give up. Time and attention count, not just money.
If you have several options and several criteria, a decision matrix helps you compare them side by side. Use it to structure thinking, not to outsource it: if the matrix picks an option that feels wrong, find out why before overruling it or trusting it.
Step 4: Gather the information that would change your mind
More information is not always better. It costs time, and past a point it mostly adds confidence, not accuracy. Gather only what could change your decision.
Ask: “What would I need to learn to switch from my current favorite to another option?” Then go find that, and stop.
Good sources, in rough order of value:
- Base rates. How do decisions like this usually turn out? How long do projects like this usually take? Outside numbers correct for optimism.
- People who have done it. Especially those for whom it went badly. Ask for the graveyard, not just the success stories.
- A small test. If you can try a cheap version of the option, real evidence beats any amount of research.
Beware of searching only for information that supports your favorite. That is confirmation bias, and it is at its strongest in this step. One fix: for every source that supports your favorite, look for one that argues against it.
Step 5: Stress-test your favorite
By now you probably have a favorite. Stress-testing means attacking it on purpose before reality does.
Two tools do most of the work:
- A pre-mortem. Imagine it is a year from now and the decision turned out badly. Write the story of why. The psychologist Gary Klein described this technique in a 2007 Harvard Business Review article; it works because it gives you permission to name risks you would otherwise wave away.
- The Bias Radar. Check the seven biases that most often distort this kind of decision. For a business decision, that means confirmation bias, the planning fallacy, overconfidence bias, survivorship bias, the sunk cost fallacy, anchoring bias and the bandwagon effect. For each, ask: “Is this pushing me toward my favorite?” The cognitive biases hub lists the Radar for career, money, relationships and business.
You can add mental models here as a Model Stack, three models for one situation. For a product launch, for example: pre-mortem, Pareto principle and second-order thinking.
If the stress-test finds a fatal flaw, go back to step 3. If it finds fixable risks, add the fixes to your plan.
Step 6: Decide and write down your prediction
Commit to the choice, and write down three things before you act:
- What you chose and why (two or three sentences).
- What you predict will happen, specifically enough that you’ll be able to tell later whether you were right.
- Your confidence, as a percentage.
This takes five minutes and is the single most useful habit in the process. It protects you from hindsight bias (the feeling that you “knew it all along”) and gives you a record to learn from. A decision journal gives this step a permanent home.
If you are still torn between close options, that is information: either choice is probably reasonable. Pick one, and set a date to revisit if the decision is reversible.
Step 7: Review on a set date
Review means comparing what happened with what you predicted, on a date you set when you decided. Most people skip this step, which is why most people’s decision making doesn’t improve with experience.
In the review, answer:
- What happened, and how close was it to my prediction?
- Was my confidence about right, too high or too low?
- How much of the outcome was my decision, and how much was luck?
- What would I do differently in the process (not just the choice)?
Separating decision quality from outcome quality is the key. A good decision can turn out badly; a bad one can turn out well. Judge the process on what you knew and could have known at the time.
Running the process faster
Seven steps sound like a lot. In practice, a DWS 4 decision might take 45 minutes: five minutes to frame, two to size, ten to map, fifteen to gather, ten to stress-test, three to write down. Only the biggest decisions need days.
If you want help keeping to the steps, the Mentarium Decision Coach runs this process inside Claude or ChatGPT, asking the framing, mapping and stress-test questions in order. The tools page has the matrix and the DWS calculator.
The Mentarium take
Most decision-making process diagrams look like a straight line that ends at “choose.” That is the part we think is backward. The choice is the least interesting step: by the time you get there, the quality of your decision has mostly been settled by how you framed it, how many options you found and how honestly you tested your favorite.
The step that makes the process compound over time is the one after the choice: writing a prediction and reviewing it. Without that, you just have opinions about your past decisions, shaped by hindsight. With it, you have data. If you are going to cut steps for speed, cut from the middle. Keep the frame, and keep the review.
FAQ
What are the 7 steps of the decision making process?
In this guide: 1) frame the decision, 2) size it with the Decision Weight Score, 3) map the options, 4) gather the information that would change your mind, 5) stress-test your favorite, 6) decide and write down your prediction, 7) review on a set date.
What is the most important step in decision making?
Framing. If you define the decision wrongly, every later step works on the wrong problem. A close second is review, because it is the only step that teaches you whether your process works.
How long should the decision making process take?
It depends on the decision. Easy-to-undo, low-impact choices deserve minutes. Permanent, high-impact choices with many unknowns deserve the full process and at least a night's sleep. The Decision Weight Score is a quick way to tell which is which.
Is the rational decision making model realistic?
Not in its pure form. The classic rational model assumes you can list every option and know every outcome, which real decisions rarely allow. Herbert Simon's idea of bounded rationality is closer to reality: you search until you find a good-enough option, within limited time and information. A practical process works with those limits.
How do groups follow a decision making process?
The same steps apply, with two additions: collect each person's view independently before discussion, so the first speaker doesn't anchor everyone, and make clear who has the final say. A group pre-mortem is especially useful in the stress-test step.