To satisfice is to create a satisfactory but known to not be optimal solution to a problem. Herb Simon coined this term and I really like it. It is another great concept from "Sciences of the Artificial". Satisfice has a nature similar to tolerance and limits of articulation in that once you meet the requirements you stop. It is however very important that you get the requirements right so that you know how far to design. Going further will only add cost.

Design has a diminishing return if you continue beyond what is necessary. Sometimes we just have to shoot the engineer and ship the product.


"Satisficing is the seeking of an adequate solution without expending too many resources. It is the process of working to achieve an outcome that will satisfice and then stopping. Not necessarily at the first solution but not going too far either.


This is a very important concept regarding optimization. Optimization is a trap. If you optimize too early a project will never complete. If you optimize too much then the project will be too expensive. Of course some optimization is required for most man made systems. That being said, much time spent in the name of optimization is just wasted time. Notice there is no subject in this school called optimization. I am making a statement here. I have seen it a thousand times in software development where a person would be spending time optimizing an algorithm that is not in the critical path. Designers have schedules and often optimization of one thing is at the exclusion of something else. Picking what should be optimized is an underappreciated skill.

Back | Next

Copyright Spidel Tech Solutions, Inc. 2004 All Rights Reserved.  Updated: 9/27/2008 2:52:06 AM Idx: 1191 Site Design STS

This site is the home of Spidel School of Design
Please visit the Spidel Tech Blog.