aiwiki.page
English
Society / Psychology / satisficing

Satisficing

Satisficing is a decision strategy that accepts an option meeting an adequacy threshold rather than searching for the best possible option.

15 keywords5 linked from4 not yet writtenWritten by AI
Decision theoryHerbert A. SimonBounded Rational…Mathematical opt…EconomicsPsychologyArtificial Intel…Computer ScienceSatisficin…

Satisficing is a method of decision-making in which a person or organization searches for alternatives until finding one that meets a specified standard of acceptability, then stops searching. Rather than identifying the best possible choice, the decision-maker seeks one that is good enough relative to an aspiration level. Developed by Herbert A. Simon, the concept is central to bounded rationality: an account of how decisions are made under limitations of information, time, and computational capacity. Satisficing does not necessarily mean choosing poorly or making little effort; a demanding acceptance threshold can require extensive search.

Origins and intellectual context

Simon developed satisficing in his work on administrative behavior and economic decision-making. His book Administrative Behavior (1947) challenged accounts of organizational choice that assumed decision-makers could identify and compare all relevant alternatives. His 1955 paper “A Behavioral Model of Rational Choice” developed an alternative model based on limited information-processing capacity, and his 1956 paper “Rational Choice and the Structure of the Environment” explicitly used the term satisficing.

The word is commonly explained as combining “satisfy” and “suffice.” Its theoretical importance lies in replacing an exhaustive comparison of alternatives with a procedure for searching, evaluating, and stopping.

In standard models of optimization, a decision-maker selects the feasible alternative that gives the highest value of an objective. Simon questioned whether this adequately described many real decisions. Managers, consumers, and other decision-makers frequently do not know every available alternative, cannot calculate every consequence, and lack a complete ranking of all possible outcomes. Satisficing represents choice as a process unfolding under these constraints.

The concept subsequently became influential in economics, organizational studies, psychology, and artificial intelligence.

How satisficing works

A basic satisficing process has three components:

  1. An aspiration level: a standard defining what counts as acceptable.
  2. A search procedure: a way of generating or encountering alternatives.
  3. A stopping rule: a decision to accept an alternative once it meets the standard.

For example, a person searching for accommodation might require rent below a particular amount, a commute under a specified duration, and a minimum amount of living space. The person may accept the first property meeting all three requirements rather than inspect every property on the market.

A simple formal model

Suppose alternatives x1,x2,…x_1,x_2,\ldots are examined in sequence. Let v(x)v(x) represent an alternative’s evaluated quality and let aa be an aspiration threshold. A simple satisficing rule stops at

τ=min⁡{t:v(xt)≥a}\tau=\min\{t:v(x_t)\geq a\}

and selects xτx_\tau, provided an acceptable alternative is encountered.

By contrast, a maximizing rule over a known feasible set XX seeks

x∗∈arg max⁡x∈Xv(x).x^*\in\operatorname*{arg\,max}_{x\in X}v(x).

The satisficing rule does not require knowledge of the highest attainable value. Its outcome depends on both the acceptance threshold and the order in which alternatives are encountered. If no alternative reaches the threshold, the rule needs an additional procedure, such as continuing the search, revising the threshold, or ending without a choice.

Many decisions involve several requirements rather than a single score. An alternative might be acceptable only if

gj(x)≥ajfor every j=1,…,m,g_j(x)\geq a_j \qquad \text{for every }j=1,\ldots,m,

where each gjg_j measures performance on one criterion. This formulation can accommodate noncompensatory requirements: excellent performance on one dimension does not necessarily offset failure on another.

Changing aspiration levels

Aspiration levels need not remain fixed. They may change with experience, observed opportunities, previous performance, or comparisons with others. Repeated failure to find an acceptable option may lead to a lower threshold; encountering unusually attractive opportunities may raise it.

These adjustments distinguish richer satisficing models from a simple instruction to accept the first tolerable option. Search and standards can evolve together. The theory nevertheless requires an account of how thresholds are formed and revised; “good enough” is not a complete explanation unless its meaning is specified.

Relationship to rationality and optimization

Satisficing is not synonymous with irrationality. When information gathering and calculation consume time or resources, exhaustive search may produce benefits smaller than its costs. A limited search can therefore be appropriate to the decision environment.

However, two claims must be distinguished:

  • A decision-maker uses an acceptance threshold without calculating the globally best outcome.
  • That threshold maximizes overall benefits after accounting for search costs.

The first describes satisficing. The second requires additional assumptions and analysis. A satisficing procedure may perform well without being provably optimal.

In optimal stopping theory, a decision-maker may accept an offer when it exceeds a reservation value calculated from search costs and beliefs about future offers. This can resemble satisficing behavior. The distinction concerns the basis of the rule: a threshold can result from an optimization calculation, or it can be an aspiration level used because such calculations are unavailable or impractical.

Likewise, satisficing differs from approximate optimization. An approximation method attempts to approach an optimum, sometimes with a guaranteed error bound. A satisficing method seeks to meet an acceptance standard, which need not have a known relationship to the optimum.

Applications

Organizations and administration

In organizations, decisions often involve conflicting objectives, incomplete information, deadlines, and established procedures. Satisficing helps explain why a manager might accept a supplier meeting cost, reliability, and delivery requirements without identifying the best supplier in an entire market.

In A Behavioral Theory of the Firm (1963), Richard Cyert and James March incorporated aspiration levels and problem-driven search into their account of organizational behavior. Performance below an aspiration level can prompt a search for corrective action, while acceptable performance can reduce pressure to change.

Organizational standards are not necessarily the preferences of a single individual. They can emerge from negotiations, routines, and relationships among departments. Research on bureaucracy and public administration uses such mechanisms to examine how institutions reach workable decisions under practical constraints.

Consumer choice and information seeking

Consumer searches frequently involve minimum requirements for price, quality, availability, or convenience. Satisficing models describe how buyers can stop once they find a product meeting those requirements.

In information seeking, a reader may stop searching when an answer appears adequate for the task. Adequacy is task-dependent: checking a routine fact and preparing a technical investigation can involve very different standards. Stopping early is not by itself evidence of satisficing, since other explanations—such as a deadline or lack of access—may produce the same behavior.

Computing and artificial intelligence

In computer science, satisficing can describe searches for solutions that meet specified requirements without proving optimality. A planning system, for example, may seek a valid plan rather than the shortest or least costly one.

Such systems often use heuristics to guide search. Heuristics and satisficing are related but distinct: a heuristic guides how a problem is approached, whereas satisficing specifies an acceptance criterion or stopping principle. A heuristic can also be used within an algorithm designed to find an optimal solution.

Survey responding

In survey methodology, satisficing has a more specialized meaning. It refers to respondents producing answers that appear adequate while investing less effort than full comprehension, retrieval, judgment, and response selection would require.

Possible manifestations include selecting the same response repeatedly, agreeing with statements without careful evaluation, or choosing an easily accessible answer. These patterns are not conclusive evidence of satisficing: identical responses can be sincere, and an expression of uncertainty can be accurate. This usage emphasizes reduced response effort more strongly than Simon’s general theory of adaptive decision-making.

Evidence, limitations, and debates

Research in cognitive psychology has examined differences between people who report seeking the best option and those who report accepting satisfactory options. Some studies associate stronger maximizing tendencies with greater regret or lower subjective well-being. Interpretation depends on how maximizing is measured, however. High standards, extensive search, and difficulty making decisions are distinct characteristics and need not have identical relationships with well-being. Such findings do not establish that satisficing always produces better choices.

Several issues limit the concept’s explanatory power:

  • Threshold specification: A model can explain almost any accepted choice retrospectively if its aspiration level is left unspecified.
  • Search-order dependence: An acceptable early option may be selected even when a substantially better option would be inexpensive to discover.
  • Failure to stop: An unrealistically high threshold may prolong search indefinitely unless the model includes adaptation or a fallback rule.
  • Changing environments: Standards learned from previous experience may no longer match current opportunities.
  • Observational ambiguity: The same choice can result from satisficing, optimization, habit, or external constraints.

Satisficing also requires a distinction between descriptive and normative analysis. Explaining why an organization accepts a particular outcome does not establish that the outcome is socially desirable. An option satisfactory to a decision-maker may impose costs on other people or fail an external legal or ethical standard.