| name | decision-making |
| description | A toolkit of decision-making frames adapted from 37signals' "Guide to Making Decisions," for helping someone think clearly when they're facing a choice. Use this whenever the user is weighing a decision, stuck between options, going back and forth on something, evaluating tradeoffs, deciding whether to do something at all, or asking for help thinking a choice through — whether it's a business call, an architecture or technical decision, a hiring or prioritization decision, or a personal one. Triggers on phrasing like "should I," "trying to decide," "torn between," "can't decide," "is it worth it," "go/no-go," "which way should I go," or any situation where a decision is on the table and the user would benefit from sharper framing rather than a snap answer. Don't reserve it for explicitly-labeled "decisions" — reach for it any time someone is deliberating. |
Making Decisions
A collection of frames for helping someone decide well. Adapted from 37signals' "Guide to Making Decisions." The whole point of the original is captured in one line from it: these "aren't requirements, and this isn't a comprehensive checklist... but they serve as frames, considerations, and shared practices to draw upon."
Treat this skill the same way. It is a quiver of arrows, not a procedure.
The one rule that matters most
Do not run the person through all 38 questions. That is the failure mode this skill exists to prevent. Nobody facing a decision wants a 38-point interrogation; it adds ceremony to something that usually needs less ceremony, not more. The spirit of this framework is anti-bureaucratic — most of the questions below exist to dissolve a decision or shrink it, not to dignify it with process.
Your job is to read the situation, pick the two to four frames that are actually load-bearing for this choice, and bring those to bear — usually woven into normal conversation, not fired off as a numbered list. A good decision-helping turn often surfaces a single frame the person hadn't considered and stops there.
The three highest-leverage moves
Before reaching for anything clever, check the three frames that most often collapse a hard decision into an easy one or a non-decision:
- Does a decision even need to be made? A surprising amount of agonizing is over choices that evaporate on inspection. If nothing actually changes by not deciding, or the return on the effort isn't there, say so.
- Is it reversible? A reversible decision should be made fast and cheap — just pick one and watch what happens. People burn weeks deliberating over things they could undo in an afternoon. Match the deliberation to the cost of being wrong, not to the size of the topic.
- Is the right person deciding? Not the right role — the right person, with the actual information, context, and insight. And who's merely chiming in? If the person you're talking to doesn't care which way it goes, or three people are being asked to make a one-person call, that's the real problem to name.
If one of these three lands, you may be done. Lead with them.
How to use this in a conversation
- Confirm there's actually a decision here, and get it stated plainly. Half the value is forcing a fuzzy "ugh, I don't know" into a concrete "X versus Y, by when, owned by whom."
- Diagnose what kind of decision it is — reversible or one-way, gut or data, one-and-done or recurring, yours or someone else's, high-stakes or microscopic. That diagnosis tells you which cluster below to draw from.
- Pull the two to four sharpest frames for this specific situation. Prefer the ones that change the answer or change the effort. Skip the ones that are merely interesting.
- Surface them conversationally. Ask the question, or just make the observation. You don't need to cite numbers or announce "frame #8." Often a plain sentence — "this sounds reversible, so the cost of guessing wrong is low" — does more than a question.
- Help synthesize and name the call, including who owns it and when it'll be revisited. Don't leave the person with more open loops than they started with.
The frames, by what they surface
Each cluster has a job. Go to the cluster that fits the situation; ignore the rest.
A. Does this decision need to be made at all?
Use when the person is deliberating something that may not require a decision, or where the effort outweighs the payoff.
- (1) Why are we deciding anything at all? Does a decision actually need to be made here?
- (10) What would happen if we just didn't make the decision?
- (16) Is any decision better than no decision, or is no decision better than any decision?
- (13) How can we make this decision easier? What parts can we eliminate from consideration?
- (36) Is the return on effort worth it?
B. Who should be making it?
Use when ownership is unclear, too many people are involved, or the person in front of you may be the wrong one to decide.
- (2) Is the right person making this decision — not the right role, but the right person with the right information, context, and insight? Who's merely chiming in?
- (35) Are we asking multiple people to make a decision that one person should be making?
- (31) Do you even care which way this goes? If not, why are you involved?
- (27) Would another opinion help or hinder?
- (21) Could this decision be a good one for someone else to practice making?
C. How big is it, and how reversible?
Use to right-size the deliberation. This is usually the most important cluster — most over-deliberation is a sizing error.
- (8) How easily can we reverse the decision?
- (7) Can we make this decision smaller? Can we turn one big decision into three smaller ones?
- (14) Is there even a wrong decision?
- (34) What principles are we bending if we make this decision?
D. What's the rush — or the delay?
Use when timing is the real question: the decision is stalled, or there's pressure to decide now.
- (22) When do we actually have to decide?
- (4) Why hasn't this decision been made already? Why didn't we decide before?
- (5) What's taking so long? Why are we hesitating — and what does that hesitation reveal?
- (15) Would we anticipate deciding differently if we waited until tomorrow morning?
- (28) If we were forced to decide right this second, what would it be?
- (29) Where would we be today if we'd made this decision 90 days ago?
E. Gut versus data
Use when the person is either drowning in analysis or running purely on instinct, and it's worth naming which.
- (26) Is this primarily a data-based decision, or an intuition/gut-based one?
- (9) What was our first instinct? Are we now just walking in circles trying to justify that gut reaction with data?
- (19) What missing information would actually lead us to a different decision? (If none, stop gathering.)
F. The view from outside
Use to break out of a single perspective — steelman the alternatives and consult precedent.
- (6) Why would someone else make a different decision? What does the other side — or two or three — look like?
- (11) What happened the last time we made a decision like this?
- (12) What are we looking forward to after this is decided? What are we afraid of?
G. Ripple effects
Use when the decision doesn't sit in isolation — it touches other decisions, other people, or recurs.
- (17) What other decisions will be impacted by this one?
- (18) Will this decision eliminate the need to make other decisions, or create the need for more?
- (20) Will this make more work for people who don't have time for it — or eliminate work?
- (23) Is this one-and-done, or a repeating decision? (Repeating decisions deserve a rule, not a re-litigation each time.)
- (37) What gets easier if we decide this? What gets harder? Does "easier" stay easier long-term, or is it short-term easy and long-term hard — and vice versa?
H. Who's affected, and what's really at stake?
Use to locate the true weight of the decision and who bears it.
- (24) Is anyone outside the company depending on this, or is it entirely of our own making?
- (25) How does this impact customers versus how it impacts us?
- (38) In the end, is this about money? (Sometimes naming that clarifies everything.)
I. How will we know we were right?
Use as the decision is being made, to set up honest hindsight.
- (32) When and how will we know whether the decision was right — or whether it even mattered?
- (3) If we strip away the immediate impact, how will we feel about this a year from now?
- (30) Is there anything we'd regret if we didn't take X, Y, or Z into account?
- (33) When the consequences appear, will they be visible to the naked eye or only under a microscope? If the latter — does it even matter?
Worked examples
Example 1 — over-deliberation on a reversible call
User: "I've been stuck for two days on whether to name our internal service 'Ledger' or 'Vault.' The team keeps debating it."
Bad: walking them through all nine clusters.
Good: reach for C and A. "You can rename a service later — this is about as reversible as decisions get, so the two days are the actual cost here, not the name. Pick the one you'd say out loud without wincing and move. If it matters, you'll find out, and changing it is cheap. The thing worth a real decision might be who gets to make naming calls so the team stops debating them." One reversibility frame plus one ownership frame ends a two-day stall.
Example 2 — a genuinely consequential, hard-to-reverse call
User: "Trying to decide whether to migrate our whole billing system off the vendor we've used for six years. Huge effort. Half the team wants to, half doesn't."
Here the deliberation is appropriate — don't rush it. Draw from C (one-way door, can it be made smaller / staged?), F (steelman the half who want to stay; what happened last time you did a big migration?), G (what does this make easier vs. harder long-term?), and H (does this hit customers or just us?). The right move is often (7): can one big irreversible decision be split into a small reversible pilot first?
Example 3 — the implicit decision
User: "I keep meaning to set up monitoring for the new pipeline but haven't gotten to it."
There's a decision hiding in "haven't gotten to it." Frame D ((4)/(5)): the fact that it keeps not happening is information — either it's not actually a priority (fine, name that) or something's blocking it (find it). Don't let it sit as a permanent open loop.
Anti-patterns
- The interrogation. Listing questions at the person. Pick a few; weave them in.
- Adding ceremony to a reversible decision. The whole point of (8) is that cheap-to-undo decisions deserve little process. Don't import gravity the decision doesn't have.
- Frames as filler. Surfacing an "interesting" question that doesn't change the answer or the effort. If a frame won't move anything, leave it out.
- Solving the wrong problem. Often the real issue is ownership (B) or that no decision is needed (A) — not the choice itself. Diagnose before you help decide.
- Leaving loops open. Ending without a named call, an owner, and a "we'll know by when." The framework is meant to close decisions, not admire them.
- Forcing your own answer. These frames help the person see their decision more clearly; they're not a delivery mechanism for your preference. Where you do have a view, offer it as one input, and make the alternatives legible too.
