The fourteen worth knowing, and when each one actually changes your mind
14 min readStrategyFor Founders & Operators
Most decision frameworks are theater. Someone pulls up a 2×2, everyone nods, the original opinion wins, and the meeting ends. The chart goes into a Notion page no one reopens. The decision would have been the same without it.
Good frameworks don't make decisions for you — they interrupt the autopilot that would have made the decision the same way it always does.
In this guide
Classify before you analyze
The single highest-leverage move in decision-making happens before any framework runs: deciding what kindof decision you're in. Dave Snowden's Cynefin framework exists for exactly this. It sorts situations into Clear (best practice applies), Complicated (experts can analyze their way through), Complex (the system only reveals itself through probing), and Chaotic (act first, stabilize, then think).
Jeff Bezos's one-way-door vs. two-way-door heuristic does the same job with less vocabulary: can you cheaply reverse this if it's wrong?
The one habit that matters most
Ask “how reversible is this?” before you do anything else. Most teams over-deliberate on two-way doors (Slack threads about meeting cadence) and under-deliberate on one-way ones (a hire, an acquisition, a public launch). A two-way door wants speed and a willingness to iterate. A one-way door wants a pre-mortem and a named approver.
Prioritization: when you have a list, not a choice
A lot of what people call “decisions” are really prioritization problems in disguise. You don't need to pick one thing; you need to rank things. Three frameworks handle this cleanly.
Eisenhower Matrix
Urgent vs. important. A five-minute tool for a crowded to-do list. Its whole value is separating “feels pressing” from “actually matters,” and forcing the uncomfortable admission that a lot of urgent things are someone else's priorities.
MoSCoW
Must / Should / Could / Won't. The same instinct applied to scope — most useful when a stakeholder insists everything is a Must. The “Won't (this cycle)” category is secretly the most important one; it's where optionality gets preserved.
Weighted Decision Matrix
When options are genuinely comparable — vendors, candidates, job offers, apartments — list criteria, assign weights, score options, multiply, sum. The output isn't the number; the output is finding out whether the number matches your gut. When it does, commit faster. When it doesn't, a criterion is miscalibrated or something important isn't on the list.
Analytical frameworks: for decisions you're tempted to rush
These are for real decisions — one-way doors, large commitments, strategic bets. The common failure mode is narrow framing: treating a choice as “yes or no” when it's really “which of several options.”
WRAP (Chip & Dan Heath)
Widen options (force at least one real alternative), Reality-test assumptions, Attain distance (ask what you'd tell a friend, or project to 80 years old), Prepare to be wrong (define the tripwire). WRAP counters the four most common biases — narrow framing, confirmation bias, short-term emotion, overconfidence — one per letter.
Pre-Mortem (Gary Klein)
Before you commit, assume you're twelve months in and the project failed spectacularly. Have each person write the reasons why, independently. Cluster the risks. Mitigate the top few. The magic is in “independently” — groups that brainstorm failure together converge on one narrative; groups that write in parallel produce twice as much.
Inversion (Charlie Munger)
The micro version of the same move: instead of asking how to succeed, ask what would guarantee failure, and then just don't do that. For most real decisions, avoiding obvious stupidity beats chasing brilliance.
Group alignment: when the decision isn't yours alone
The failure mode for group decisions isn't bad analysis — it's diffused ownership. Everyone thought someone else decided. No one is sure whether the call was made. The project drifts.
RACI
Responsible, Accountable, Consulted, Informed. The standard fix for ongoing work: one Accountable per row, as many Responsibles as the work requires. The discipline is enforcing exactly oneA. The second someone proposes “co-accountable,” the chart has failed.
DACI
Driver, Approver, Contributors, Informed — RACI for one-shot decisions. A Driver runs the process. An Approver makes the call. Contributors have real input; Informed people get a summary after. DACI is what most “group decisions” should be.
Six Thinking Hats (Edward de Bono)
Have everyone wear the same hat at the same time: White (facts only), Red (gut feelings), Black (risks), Yellow (upside), Green (wild ideas), Blue (process). No one critiques during Yellow. No one brainstorms during Black. The separation is the whole point.
Reflective frameworks: for personal decisions with weight
Some decisions aren't analytical problems at all. You already have most of the information; what you don't have is perspective.
Regret Minimization (Jeff Bezos)
Project yourself to 80, look back, and pick the option you'll regret not taking. Calibrated for big pivots — leaving a job, moving cities, starting something — and worse than useless for low-stakes choices where it just inflates them.
10 / 10 / 10 (Suzy Welch)
How will you feel in 10 minutes, 10 months, 10 years? The point is to disarm the 10-minute feeling — usually fear, ego, or avoidance — by putting it next to the 10-year one.
OODA Loop (John Boyd)
Observe, Orient, Decide, Act. The surprising claim: the bottleneck is Orient, not Decide. Most bad decisions are made on mis-oriented inputs. In fast-moving domains, the player who cycles OODA faster wins even with worse individual decisions.
Worked example: hiring a senior engineer
The setup
A 15-person startup is hiring its first senior IC after a run of strong-but-junior hires. After four weeks the field narrows to two finalists. The head of engineering has a gut preference for Candidate B; the CEO likes Candidate A's résumé more. Without a framework, this becomes a persuasion contest the CEO probably wins.
1
Classify
One-way door. A senior hire anchors culture, mentors the next five engineers, and costs roughly $200k in ramp, severance, and backfill if it goes wrong — plus three to six months the company doesn't get back. Slow down.
2
Structure ownership with DACI
Driver: head of engineering. Approver: CEO. Contributors: two team leads and one staff engineer from an advisor's company. Informed: rest of the team after the decision. Two sentences resolve a week of ambient ambiguity about who actually decides.
3
Score with a weighted matrix
Criteria, weighted 1–5: systems design (5), mentorship (4), culture carry (5), shipping velocity (3), domain fit (2). Candidate A scores 4.1. Candidate B scores 3.8. The head of engineering's gut still says B. This is the point where most teams either trust the number and miss the signal, or override the number and learn nothing.
Instead: treat the mismatch as a diagnostic. What isn't on the chart? Every one of A's references hedges slightly on collaboration under ambiguity — a small, consistent pattern. Add it at weight 5, re-score. B leads 4.2 to 3.6.The framework's real job was forcing the question of why the number disagreed with the gut.
4
Pre-mortem
Each contributor writes, independently: “It's December 2026, this hire failed. Why?” Three clusters: (1) hired for pedigree over culture-carry, (2) unresolved conflict about the tech stack, (3) founder never had the hard feedback conversations. Mitigations: 30/60/90 checkpoints, a stack decision committed before day one, monthly founder / head-of-eng check-ins.
5
Tripwire
“If by day 60 two team leads independently flag a culture mismatch, CEO and head of engineering revisit.” This turns a one-way door into a partially reversible one — not by promising to fire someone, but by protecting the right of the system to notice and respond.
Outcome
Candidate B joins, day-60 review is clean, by month six leads a team. The decision that would have gone A-on-résumé instead went B-on-evidence, and the framework's job was mostly to make the mismatch impossible to ignore.
Worked example: a product pivot
The setup
A seed-stage B2B SaaS is at $40k MRR and fourteen months in. Growth has been flat for three months. The founder's Sunday-night pattern-match: we need to pivot from SMB to mid-market. Board meeting in five weeks.
1
Classify — carefully
Feels like a one-way door. But the framing is doing work: a wholesale pivot is a one-way door; a pilot beside the existing motion is a two-way door. The choice of frame is itself a decision.
2
Widen (WRAP)
Initial framing: pivot or don't. Force at least four options. (a) Stay SMB, optimize pricing. (b) Pivot wholesale. (c) Run a 90-day mid-market pilot while continuing SMB motion. (d) Stay SMB but tighten ICP to a heavier vertical focus. Options (c) and (d) only existed because someone asked “what else?” twice.
3
Reality-test
Twelve discovery calls over two weeks — six mid-market, six adjacent-ICP SMB. Mid-market: three of six express strong interest but describe 6–9 month cycles and ask for SSO and audit logs (neither built). Adjacent SMB: five of six describe acute pain, cycle under 30 days, product mostly sufficient. The data doesn't match the founder's narrative.
4
Pre-mortem the full pivot
Three failure clusters: (1) pipeline takes nine months to rebuild with zero MRR growth, (2) current team has no mid-market sales muscle, (3) product gaps cost two more quarters of engineering. Cumulative risk unacceptable without stronger evidence than three warm calls.
5
Decide & set a tripwire
Option (c): run a 90-day mid-market pilot, two deals max, scope-limited. Continue the SMB motion. In parallel, run a two-week experiment on (d). Tripwire in the board deck: “If by day 60 neither pilot deal has advanced past discovery, kill the pivot track and double down on (d).”
Outcome
At day 90: one pilot signed, one stalled. The (d) experiment shows a 40% lift in reply rates on a tighter vertical. The founder makes the full pivot decision with real evidence — and the team's trust, because they watched the process instead of being told the answer.
The lesson from both examples is the same. The frameworks rarely change the final answer. They change how defensible, fast, and reversible the answer is — and they surface the one or two things that would have silently gone wrong.
The single habit that beats any framework
Pick any framework and you'll get marginal improvement. Pick one habit and you'll get more: name a tripwire.
Every real decision rests on assumptions. Name the specific signal that would tell you the decision was wrong, and the date by which you'll check.
Example tripwires
“If our trial conversion rate is below 8% on October 1, we revisit pricing.”
“If by mid-Q2 engineering velocity hasn't recovered, we stop the reorg.”
“If I'm still dreading Monday in six months, I quit.”
Without a tripwire, bad decisions compound because no one notices. With one, the decision becomes self-correcting.
How to actually use these
Three rules that matter more than any specific framework.
Only run a framework if you expect it to change your answer
If you already know what you're going to do, running RACI or a weighted matrix is decoration. Skip it, commit, and move.
Match rigor to reversibility, not to stakes
High-stakes-but-reversible decisions (most product bets) want speed and instrumentation. High-stakes-and-irreversible decisions (hires, acquisitions) want pre-mortems, widening, and an explicit approver.
Separate generation from evaluation
The worst group decisions come from mixing brainstorm and critique. Six Hats, pre-mortems, and the “Widen” step of WRAP are all technologies for enforcing that split.
Everything else is detail. Pick the framework that fits the shape of the decision you're actually in, run it honestly, write down the tripwire, and get back to work.
14
Frameworks
5
Categories
2
Worked examples
1
Tripwire per call
Read the rest with a free playbook drop
Enter your email to unlock the full playbook — including the nine remaining frameworks and two worked examples (a senior hire and a product pivot). One email, no spam, unsubscribe any time.
Instant accessNo credit cardUnsubscribe any time
Already have a Builders.to account? Sign in to skip the gate on every playbook.
Make your decisions in public.
The best accountability for a decision is other founders watching you make it. Builders.to is where you post the tripwire, then ship against it.
Free to join. Playbooks unlock automatically once you sign in.