Satisficing
Satisficing: Define the minimum acceptable outcome in advance. Choose the first option that meets it, then stop searching. Optimization is rarely worth its cost; a good decision made quickly is usually better than an optimal decision made slowly.
What Is Satisficing?
Herbert Simon, winner of the 1978 Nobel Prize in Economics, introduced the concept of satisficing in a 1956 paper as part of his broader theory of bounded rationality. Classical economics assumed that decision-makers had unlimited information, unlimited cognitive capacity, and unlimited time — and would therefore always optimize. Simon observed that real decision-makers have none of these. They face uncertainty, cognitive limits, and time constraints. Given these constraints, optimizing is often impossible and always costly.
His alternative: satisficing. Rather than defining the best possible outcome and searching until you find it, you define a "good enough" threshold and accept the first alternative that meets it. The threshold should be set high enough to ensure a genuinely good outcome, but not so high that you're effectively still optimizing.
The insight: the cost of continued search — time, cognitive effort, decision fatigue, missed opportunities from delay — is often higher than the expected value of finding a marginally better option. For most decisions, the difference between "good" and "optimal" is small; the difference between "fast" and "slow" is large.
Satisficing is not the same as settling. Settling implies accepting a poor outcome. Satisficing implies setting a rigorous standard and accepting the first option that genuinely meets it. The discipline is in setting the threshold correctly — high enough to matter, not so high as to require indefinite search.
How It Works
Step 1: Define your acceptability threshold before searching
"I need a software vendor who: (1) can handle our data volume,
(2) has SOC 2 certification, (3) costs under $5K/month,
(4) has references from companies similar to ours,
(5) can deploy within 90 days."
This is the threshold. Do not adjust it during the search
to fit a candidate you've become attached to.
Step 2: Search through alternatives
Evaluate each option against the threshold as you encounter
it. This is sequential search, not simultaneous comparison.
Step 3: Accept the first alternative that meets the threshold
Stop searching. Do not continue looking for something better.
The opportunity cost of continued search is real.
Step 4: If no alternatives meet the threshold...
Lower the threshold (if one requirement is disproportionately
constraining and less important than the others).
OR: accept that no good option currently exists.
Do not: lower the threshold to justify accepting an option
you've already encountered that doesn't meet it.
Real-World Examples
Example 1: Job Offers
A software engineer is job searching. She has identified her threshold: a company with interesting technical problems, total compensation above $160K, strong engineering culture, and a role that offers leadership opportunity. She is interviewing at six companies.
Satisficing approach: when she receives an offer from Company C at $175K with strong engineering culture, interesting problems, and a lead role, she accepts — even though Company D and Company F are still in her pipeline.
Why? The cost of continuing the search (time out of market, relationships with companies already passed, the possibility that Company C reschedds) may be higher than the marginal improvement a Company D offer might provide. Her threshold was met; she accepts.
The contrast: an optimizing approach would continue the process to completion, compare all six offers simultaneously, and select the maximum — at the cost of several additional weeks of uncertainty and the risk that other processes collapse.
Example 2: Netflix's Recommendation Problem
Netflix has described its recommendation engine's approach as a satisficing problem, not an optimization problem. The goal is not to surface the best possible film for a user at this moment — that would require solving an intractable optimization problem across millions of options and billions of signals. Instead, the system identifies films that meet a threshold: the user is likely to enjoy it, is likely to watch more than 70% of it, and is likely to give it a positive rating.
The first satisficing recommendation that meets this threshold is surfaced first. This produces a good-enough outcome (the user watches something they enjoy) far faster than optimization would.
Example 3: Hiring a Contractor
A startup needs to hire a freelance developer for a 3-month project. The options for finding candidates: a personal network referral, a job board post, a recruiting agency, LinkedIn outreach. The threshold: comfortable working with the technology stack, available to start within two weeks, hourly rate under $150, strong references from one recent client.
The founder posts on the job board and receives 20 applications. On review, Candidate 4 meets all threshold criteria. Satisficing says: interview candidate 4 seriously; if they pass a technical screen and reference check, make the offer. Don't wait to review all 20 candidates.
The risk of satisficing here: a candidate ranked 1 or 2 by quality might be better. The benefit: the project starts weeks earlier, which matters for a time-critical deliverable. The satisficing threshold was set correctly; the expected cost of the marginally better candidate is outweighed by the expected benefit of faster start.
When to Use It
✅ For decisions where good-enough is genuinely adequate and the marginal benefit of optimal is small.
✅ When speed matters — when a faster good decision beats a slower optimal decision.
✅ For repeated, similar decisions where the cost of optimizing every instance is high and the cumulative benefit of satisficing is large.
✅ When decision fatigue is a real risk — satisficing conserves cognitive resources for decisions where optimization is genuinely important.
❌ For irreversible, consequential decisions where the marginal cost of the non-optimal choice is very high. Choosing a medical treatment or a business partner may warrant optimization.
❌ When the options haven't been sufficiently explored to set a meaningful threshold. If you haven't looked at enough options to know what "good enough" looks like, set a time-limited exploratory phase before applying satisficing.
Model Combinations:
| Combine with | Effect |
|---|---|
| Two-Way Door | Satisficing is the right decision process for Two-Way Door decisions; optimization for One-Way Doors |
| Decision Matrix | Use a matrix to set the threshold; satisfice against it |
| Eisenhower Matrix | Apply satisficing to the "urgent, not important" quadrant; reserve optimization for "important" decisions |
Common Misuses and Limitations
Misuse 1: Setting the threshold after seeing the options. The threshold must be set before the search. If you adjust it to fit a candidate you've already seen, you're rationalizing, not satisficing.
Misuse 2: Confusing satisficing with settling. If the threshold is too low, satisficing produces consistently mediocre outcomes. The quality of satisficing is determined by the quality of the threshold.
Misuse 3: Applying satisficing to decisions where optimization is worth the cost. For high-stakes, irreversible decisions, the marginal benefit of optimal may be large enough to justify the search cost.
Related Models
Two-Way Door: Satisficing is the natural decision process for Two-Way Door decisions, which can be corrected quickly.
Opportunity Cost: The opportunity cost of continued search is the key consideration in determining when to satisfice.
FAQ
How do I set the right threshold for satisficing?
The threshold should reflect the minimum outcome you'd genuinely be satisfied with — not your ideal, but the floor below which you'd regret the choice. A practical approach: write down your threshold before you see any options, based on your needs and the realistic alternatives available. Then stress-test it: if the only available options are at the threshold, would you proceed or continue searching? If you'd continue searching, the threshold is too low.
Doesn't satisficing always produce worse outcomes than optimizing?
No — and this is Simon's key insight. Optimization has costs: time, cognitive effort, opportunity cost of delay, decision fatigue, and the risk of losing a good option while searching for a better one. When these costs are included, satisficing often produces better overall outcomes than optimization. The conditions where optimization wins: decisions with very large differences between the best and good-enough options, decisions where timing doesn't matter, and decisions where the search cost is genuinely low.
What is the best resource for learning more about satisficing?
Herbert Simon's original paper 'A Behavioral Model of Rational Choice' (Quarterly Journal of Economics, 1955) is the foundational source. Barry Schwartz's The Paradox of Choice (2004) applies the concept to consumer decisions and describes research on maximizers (optimizers) vs. satisficers, finding that satisficers tend to be happier with their choices despite often getting objectively lower-ranked outcomes.
Apply This Model with AI
Describe the decision you're trying to make and the options you're considering in MindMax. The AI will help you define an explicit acceptability threshold and evaluate your options against it rather than trying to identify the perfect choice.
🚀 Apply Satisficing in MindMax →
Further Reading
- Herbert Simon, "A Behavioral Model of Rational Choice," Quarterly Journal of Economics (1955) — The foundational paper; technical but readable.
- Barry Schwartz, The Paradox of Choice (2004) — The accessible treatment of satisficing vs. maximizing in everyday decisions.
This page is part of the MindMax Mental Models Knowledge Base.