A decision-making framework is a repeatable way to structure a choice: what kind of problem it is, who decides, which options and criteria count, and how much time and evidence the decision deserves. The best-known ones answer different questions, from Herbert Simon’s satisficing (accept a good-enough option when search is costly) to Cynefin (match the method to the type of problem) and Amazon’s one-way and two-way doors (match the process to how reversible the decision is).
How to choose the right framework
Where you see it today
A preregistered experiment with 758 knowledge workers, run with Boston Consulting Group, found that those using GPT-4 completed 12.2% more tasks, 25.1% faster and at higher quality on tasks inside the AI’s frontier, but were 19% less likely to reach a correct answer on a task outside it (Dell’Acqua et al., 2023). The first decision is whether a task is one AI handles well.
AI tools make many product and content decisions easier to test and reverse, which moves more choices into Bezos’s Type 2 category. The risk he named also grows: using a heavyweight Type 1 process on decisions that could simply be tried.
Deploying AI agents in a workflow is often a problem whose outcomes are hard to predict in advance. Cynefin’s point that few solutions are context-free argues for small, safe-to-fail pilots in each team rather than one company-wide plan.
As companies let software act on their behalf, role frameworks such as RAPID and DACI can be extended to say which decisions an AI may recommend, which it may make, and who approves its output.
Common decision frameworks and when to use them
How to use it
- Before choosing an option, write one line on the type of problem and whether the decision is reversible.
- Give every important decision a single named decider and list who recommends and who must agree.
- For reversible decisions, set a short deadline and decide with about the evidence you have.
- For irreversible ones, write down the options, the criteria and what would change your mind.
- When AI helps, check first whether the task is one the model handles reliably, and keep a human check on tasks outside that frontier.
AI lowers the cost of analysis but not the cost of a wrong, irreversible call, so the scarce skill is classifying decisions and assigning who owns them. The Business Engineer view: companies that sort decisions by reversibility and push two-way doors down to small teams with AI support will move faster than those that run every choice through one heavy process.
Read: The Four Resolutions Decision-Making Framework →Explore everything we publish →Related: Cynefin framework · OODA loop · Satisficing · RAPID framework · Decision matrix
Frequently asked questions
What is a decision-making framework?
A repeatable structure for making a choice: it clarifies the type of problem, who decides, which options and criteria count, and how much time the decision deserves.
Which decision framework is best?
None fits every case. Use Cynefin to classify the problem, one-way and two-way doors to size the process, RAPID or DACI to assign roles, and a weighted matrix to compare options.
What are one-way and two-way door decisions?
Terms from Jeff Bezos's 2015 Amazon shareholder letter: one-way doors are consequential and nearly irreversible and need careful deliberation; two-way doors are reversible and should be made quickly.
Does AI make better decisions?
It depends on the task. In a study of knowledge workers using GPT-4, AI improved speed and quality on tasks inside its capabilities but made correct answers less likely on a task outside them.
Sources
- Simon (1955), A Behavioral Model of Rational Choice, Quarterly Journal of Economics
- The Royal Swedish Academy of Sciences (1978), Prize in Economics press release: Herbert Simon
- Snowden et al., Cynefin (The Cynefin Co wiki)
- Bezos (2016), 2015 Letter to Shareholders, Amazon.com (Exhibit 99.1 to Form 8-K), SEC
- Dell'Acqua et al. (2023), Navigating the Jagged Technological Frontier, Harvard Business School Working Paper 24-013
The full guide
The complete explainer, cleaned up and reformatted for mobile.
Satisficing

RAPID Framework

Foursquare Protocol

DACI Decision-Making

Lightning Decision Jam

Multi-Criteria Analysis

Cynefin Framework

SWOT Analysis

Personal SWOT Analysis

Pareto Analysis

Failure Mode And Effects Analysis

Blindspot Analysis

Comparable Company Analysis

Cost-Benefit Analysis

Agile Business Analysis

SOAR Analysis

STEEPLE Analysis

Pestel Analysis

DESTEP Analysis

Paired Comparison Analysis

Hickam’s Dictum

Occam’s Razor

Occam’s Broom

Outcome Bias

Principle-Agent Problem

TDODAR Decision Model

Mendelow Stakeholder Matrix


Go/No-Go Decision Making

OODA Loop










