Clashing Titans: Simple or Sneaky?
Since entering the sphere of Microsoft development, I have frequented Jeff Atwood's Coding Horror blog. I have linked to it before on this blog and again this past week, I found a post worth commenting on. Reducing User Interface Friction is Jeff's take on a post, Three Hypotheses of Human Interface Design, by Tantak Celik, Chief Technologist at Technorati.
Specifically, Jeff focuses on "cognitive load" in user interfaces. As usual, the primary sources are the best place to go for an explanation, but the 10 second summary is that the lower the cognitive load, the better the interface. The way to lower cognitive load is to decrease the number of actions, such as clicks that a user needs to take to accomplish his or her task.
While Tantak contrasts IM to email, Jeff returns to Google's ubiquitous single text box design as the kind of holy grail of user interface. So far so good for me, but that got me thinking about design, so I strolled over to Don Norman's website to read up on some of his thoughts on the topic.*
A former VP at Apple, Don is also a prolific writer about human computer interaction and design in general. So, it was a happy coincidence to see that he had also written on the subject of Google's interface. The title does a pretty good job of summarizing his thoughts, The truth about Google's so-called "simplicity". For Don, Google is not really that great of an interface because you can only do one thing with that one text box, which is search. So, MSN and Yahoo!, while their main pages look crowded are actually easier to use because they don't hide their functionality.
So, as usual, anything worth debating has strong proponents on both sides. Is Google's interface simple and wonderful or sneakily bland, a facade of simplicity over a tangle of complexity?
I go with strong and wonderful. The problem with Don's essay is that he is comparing apples to oranges. MSN and Yahoo! are web portals first and foremost that happen to have search engines. Google is a search engine that happens to have enough functionality to compare to a web portal.
The difference is easy to spot; if you go to MSN.com and want to get directions somewhere, you see a whole slew of links, that if you read through, you can spot Maps & Directions, click on that and you're on your way. On Google, you have a text box, so what do you do? You type in the address you want to go to and the first result of the search is a link to a map.
So by default, unless you just give up and don't interact with Google, you find your way to what you want. If you try the same interaction on MSN.com, you can click on the the Maps tab on the toolbar in the search results, but again, that's an extra click you have to find amongst several options.
Google is certainly not perfect in this regard, as Don rightly points out. Not everything is linked up like it should be. At one point, I remember seeing Google Groups results included as the last item in search results, which was very useful when researching coding issues, and now I don't see them anymore. Similarly, the integration with Froogle is only OK, a price for the item I want to buy is often the first link returned, but it could be more obvious.
This circles back to the original point. Google.com is primarily a search engine (really, an ad company), which has been enriched enough to often function as a portal, though not perfectly. So comparing it to MSN and Yahoo! is not really fruitful because you do not have a common baseline.
Finally, while Don derides the query language extras of Google, such as getting a definition for a word using "define:word" syntax, I love them. In effect, I get rewarded for repeated interactions with Google by getting access to a more powerful interface. While I could spend the time to setup a personalized page with MSN or Yahoo!, I prefer the approach of starting simple and building up, instead of starting with the complexity of a full grown portal ecosystem and having to decide what to bushwhack.
In the end, despite the possibility of an unfair comparison, user interface questions like these always come down to the user: what they want is what they want, no matter what you what you may want them to want.^ So, only market research and user testing can be the final arbiters of this debate. Although, with the trajectory of Google's stock, I would think that at least the Wall St. user's group has made its preference clear.
* I had the pleasure of taking a class or two about design that Don taught at Northwestern while I was still in school. His class was very informative, although a bit of an ego check. He firmly did not believe in coddling his students.
^ Please rapidly repeat this sentence ten times.