Balance satisficing and optimizing

Balance satisficing and optimizing

Aug 25, 2010



[IMAGE MISSING: 3.71-Satisfycing-vs.-optimizing-e1282776612572.jpg]
Whether looking for the root causes of your problem or identifying solutions, you must be collectively exhaustive (CE). That means looking for all the possible solutions. That’s fine in theory but in practice, that induces paralysis by analysis. So you need to learn to balance your quest for collective exhaustiveness with practical considerations.


Know the difference between satisficing and optimizing

Satisficing is a portmanteau between satisfying and sufficing. It means that you’re looking for the first adequate solution to your problem. Once you’ve identified it, you stop.Optimizing, in this context, means looking for the best possible solution. No matter how good the solutions you’ve found are, you are looking for a better one.

[IMAGE MISSING: 3.68-Satisficing-optimizing1-e1282777747100.png]
A skilled problem solver must balance satisficing and optimizing as neither extreme is beneficial.

So, applied to culinary tastes, a pure satisficer will always go to the same restaurant and order the same dish. After all, if he likes it, why take a chance? Conversely, the pure optimizer will always try a new restaurant and a new dish, because, no matter what he has tried before, surely there is something better out there.

So which one is the right approach? Neither of course. It would be too simple.


First, strive to optimize

The pure satisficer throws away innovation: he’s found one good-enough solution and will stick with it. So there’s no room for progress. On the other hand, the pure optimizer throws away real-world considerations, such as deadlines and implementation.

So neither extreme of the spectrum is the right solution. Instead, you need to be somewhere in the middle. At the beginning of your solution process, push for optimization. That’s the divergent thinking part. That’s where you’re looking for innovative ways to answer your key question. Actively looking for new answers, even absurd ones, will get you out of your comfort zone and force you to explore new ways.



[IMAGE MISSING: satisfice-2.png]
First, optimize so as to get away from your comfort zoneWe’ve talked about it in another post (and we’ll talk about it some more in future creativity-related posts): go too far to the absurd side before restraining yourself. Don’t stick with the usual suspects for solutions/explanations: go look for the irrational ones, the bad ideas, the suggestions that will make people laugh at you. In short, give it your best effort to be truly collectively exhaustive.

[IMAGE MISSING: 3.70-Dont-use-other1-e1282777816656.png]
Using "other" in the branches of an issue tree is an example of satisficing. Usually, this is not desirable.

Logic trees are very helpful to push yourself in that direction. As a general rule, don’t settle for a branch that says “others” in a tree. Instead, make a conscious effort to name the elements of that branch. For example, going back to the taylor house case, my intellectual laziness drives me to identify just a couple of factors that can render a shopping experience exceptional. “You’ve done enough”, it goes, “by identifying the concept you’ve added your value, now you can relax.” But not settling for “other” will make me think about concrete ways to achieve my original purpose. My intellectual laziness is wrong: identifying the concept does indeed add value but identifying the various concrete ways in which concept can articulate itself also adds value, perhaps even more.
Decide when to stopEventually you’ll have to decide that your analysis is MECE-enough and wrap it up. Sorry, but I don’t have a clear rule when that moment comes (no “you need to generate exactly 39 potential solutions for your problem”) because it is so case-dependent. If you’re on a deadline, you can setup a maximum time for your analysis. But beware: time pressure does not exactly support creativity, so if you consider the clock too much you might throw away good ideas. Alternatively, you can set a goal for a total number of ideas, but this is even riskier because you might sacrifice significantly quality just to reach your goal, so, if you’re using this, consider setting a ridiculously high objective and be prepared to decide which ideas actually count as a new contribution and which aren’t just quite at that level.Personally, I’m having a hard time being CE in one go. So I’ll look for solutions during, say, 1/2 hour, and then I’ll go do something else; the more unrelated the activity, the better (I favor swimming or cycling). Then I’ll get back to it after an hour or, ideally, a day and do another 1/2 hours session. I’ll iterate a couple of times by myself and then I’ll ask someone else to look for ideas on her own, at first without sharing what I’ve come up with. Then, if we still have time, we’ll compare our ideas. The point is that I really want to feel that I’ve pushed it as hard as I could before I give up.The more problems you solve, the more self-aware you become so it will get better to judge when to give up on your quest to pure MECEness.
Learn moreTwo books helped me think about these concepts. The first is The Paradox of Choice by Barry Schwartz; it introduces the concept of satisfycing versus optimizing (chapter 4 is most relevant). The second is The Contrarian Guide to Leadership by Steven Sample. Sample was the president of USC when he wrote the book. If I remember correctly, in a passage he was writing about his former life as an engineer, when he came up with an idea to power washing machines with a new mechanism inspired by his thinking of orbits of stars and how these could help his mechanical design. I thought that was the perfect illustration of applied divergent thinking.