Decision-making and strategy
Satisficing versus maximizing
Decide in advance what good enough looks like, take the first option that clears it, and stop searching.
Herbert Simon, who used the word "satisfice" (a blend of satisfy and suffice) in "Rational choice and the structure of the environment", Psychological Review, 1956, arguing that organisms adapt well enough to satisfice and do not, in general, optimize. Barry Schwartz and colleagues turned the distinction into a personality scale in 2002.
A maximizer looks for the best option, which in principle means checking all of them. A satisficer sets a threshold, which Simon called an aspiration level, and takes the first option that clears it. Simon's argument was that real decision makers have limited time, information and computing power, so they cannot maximize anyway. Satisficing is what sensible behaviour looks like under those limits, and the threshold can move up or down as you learn what is available.
It works because search has a cost and the returns fall quickly. The gap between the best option and the fifth best is often smaller than the cost of finding out which is which. Satisficing also ends the decision. You stop comparing, and you spend less time afterwards dwelling on the options you did not take.
Satisficing does not mean taking anything. A threshold set too low is just carelessness, and for one-way decisions where the options differ a lot (a cofounder, the database a system will live on for ten years) a longer search pays for itself. The common misuse is saying "good enough" without ever writing down what good enough means, which turns the method into an excuse.
Example
In November 2001 Sheena Iyengar, Rachael Wells and Barry Schwartz surveyed 548 graduating students at 11 colleges and universities as they began looking for jobs, and followed them to May 2002. Students who scored high on maximizing found jobs with starting salaries about 20% higher: a mean of $44,515 against $37,085 for those below the median. Yet the maximizers were less satisfied with the jobs they took and felt worse throughout the search. The 2006 paper in Psychological Science was titled "Doing Better but Feeling Worse".
Where it applies
When choosing a library, a vendor or a model for a task, write the acceptance criteria first (latency, cost per call, licence, a pass rate on your own test set) and pick the first candidate that meets them. Keep the exhaustive comparison for choices that are expensive to reverse. In code, this is the difference between an early-exit search and a full scan.
Think about it
Which decision are you still researching that already has an option good enough to take?
Applied it to something real? Log it.
Test yourselfWhy can satisficing be the more rational strategy, even though maximizing aims for a better outcome?
Searching costs time and attention, and the gain from finding the very best option is often smaller than that cost. Maximizers can also end up less satisfied with better outcomes because they keep comparing.
Related models
- The opportunity cost of timeThe real price of an hour is the best thing you could have done with it instead.
- One-way and two-way doorsMake reversible decisions quickly and save slow, careful deliberation for the ones you cannot undo.
- Regret minimizationFor a big personal choice, imagine yourself at 80 looking back, and pick the option you would be least likely to regret.
- Opportunity costThe cost of any choice is the net value of the best alternative you gave up, whether or not money changed hands.