"Count what is countable. Measure what is measureable. What is not measureable, make measureable." -- Galileo

Friday, November 20, 2009

Decisions, Decisions...

Looking at CMS Matrix tonight I find an astonishing 1099 listings. How can anyone make a sane decision with that many choices? This ties in nicely with Barry Schwartz's thoughts in a TED video on the paradox of choice: too many choices lead to decision-making paralysis and dissatisfaction. He provides examples with salad dressings, jeans, and cell phones, among several others. To over-simplify his thesis, too many potential choices increase the difficulty of decision-making and heighten expectations. This in turn dilutes satisfaction with one's ultimate decision.

In his book "The Paradox of Choice" he points out that even infinite choice isn't a bad thing if it can be conceptualized in a simple way, for example, in a one-dimensional matrix comparing a simple feature set. But as the dimensions of comparison go up, the computation needed to calculate an optimal choice becomes increasingly difficult. I see this at my day-job where they have been unable to decide on a corporate standard for CMS. In the end, people look for surrogates for quality to simplify the decision-making process, things like the recent Packt Awards or the Idealware report.

Another solution is to follow the herd and make the same decision others before you have made. This can easily lead to an information cascade and sub-optimal results. I've discussed these beasts elsewhere, but want to point out again that herd behavior is fragile--a little correct contrary knowledge can quickly dispel an information cascade.

Of course, one problem is that one size (CMS) does not fit all. In my case workflow and security trump everything else, so the choice is dramatically limited... and simplified. I don't spend my time agonizing about if I made the right choice or second-guessing myself.

Along those lines, in a bookstore tonight I saw a copy of "Blink," also about decision-making. After only a short skim, I'd say that I agree with the author: we often make an optimal decision within seconds but then our slower conscious analytical processes talk ourselves out of it. (This is not to say that reflective decision-making is always a bad idea, just that we should weigh our initial impressions more heavily in the reflective process.)

I see this all the time grading tests at college--the initial correct answer is erased and an incorrect answer finally selected. I tell my students all the time to not second-guess themselves unless they have a good reason.

This resonates with a number of usability studies that indicate that we make web choices within seconds or fractions of seconds. This article by Jakob Neilson has a good discussion and references on this topic. BTW, almost 60% of visits to Plone.org are less than 10 seconds. On the other hand, about 30% of the visitors stay 3-30 minutes. My guess is that Plone.org visitors aren't generally making a "use Plone" decision in 10 seconds.

Now here's an experiment for you: go to Plone.org and count off 10 seconds while you look at the site. What did you remember? Branch with bud, Plone South America, "powerful, flexible, easy to install, use, and extend."

Do the same for Drupal.org: two simultaneous releases, "critical security vulnerabilities," Whitehouse.gov.

Try OpenOffice.org: productivity suite, learn, download, help, do more, participate. Hmm? Simplifying choice, but letting the user make an informed decision about what to do next. Maybe Schwartz is on to something...

3 comments:

Matt Hamilton said...

Karl,
Interestingly, I was pointed at this site last week: http://fivesecondtest.com/ which does pretty much what you are talking about (and does it twice as fast! ;) ).

I'm certainly interested in running a few test on some of our sites on it.

-Matt

Schlepp said...

Thanks, Matt. 5-sec test is pretty cool. I'll be using it as well.

Dylan Jay said...

awesome post.

also drupal gets two dips at giving a message.

drupal.com which "advertises" drupal with portofilos etc. Message: drupal is slick, and used a lot.
drupal.org which is about engaging the community and getting hobiest to download and try it.
Plone.net doesn't really count in this regard as its not so much a plone advertising medium as a integrator advertising medium.