Googlers Advised To Limit Choice
Barry Schwartz (no, not that Barry) gave a talk at Google about his book “The Paradox of Choice“, in which he explains that, given too many choices, most people paralyze and do nothing, and need an “agent” to show them the way. You can watch the talk on Google Video, or just rely on Mahlon at BuyGoogle to explain it for you.
While the theory is sound (and no doubt Mr. Schwartz is smarter than me), it is exactly the sort of thinking I hate. Maybe most people don’t function well with that much on their plate, but I know I’m not the only one that actually enjoys examining situations from every angle to choose the most satisfactory solution. When I buy something, it’s only after I’ve tried every option, because it is just worth it knowing that whatever I purchased, I did everything in my power to validate my choice.
I hate software development that takes options away from the user. I hate that Steve Jobs seems to think the best way is to do one thing right. I don’t want one thing, I want everything. I’d rather stick with my less stable Windows computer, since my options are nigh-limitless. With all the work we put into our computer systems, we deserve better than being told what to do.
If Google wants to design software and services easy enough for the “paralyzed”, more power to them. Those people need the simpler, right-to-the-point interfaces, and Google will make lots of money serving them. Just don’t subscribe to the “we’re all fish” theory, because some of us are sharks, and we’re hungry for much more.
Actually, what Google seems to do in my opinion is better than the “least options” and “tons of options”. Google starts out with the most significant features first, and adds more features based on user feedback. The downside to that approach is that sometimes it results in initial disappointment; what Google tries to do and have been successful in some cases is make the initial features or product release so significant that users are willing to wait for the requested additions(examples Gmail, Maps, Calendar). They try to excite the user with the initial release; the only problem is with some initial release sthe “significant” feature was not that exciting to many users(e.g. Google Base, Gtalk) Still, it’s a good balance; like in the case of Gtalk the new features trickling in keeps Gtalk users satisfied. And it keeps a sense of loyalty for google users because you see your requests being implemented.
Comment by or | May 8, 2006
You didn´t get it. He said, that everybody wants the best capability of stuff, but finally they don´t enjoy it. Look at Gmail and new Yahoo! Mail. That´s exactly what he says. Less choices (Gmail) means better usability than Yahoo!. When 99.9% people never use some smart fuctionality, why bother them with it? And even more - webdesigner should know better, how people use things. Let them choosing and they will go elsewhere. Another example - personalized homepages. Do you know, what you want on your homepage? Yes - mail, RSS reader… what else? So you can choose some of hundreds modules. But that´s really diffucult. Don´t force people to think.
Comment by Jim | May 8, 2006
Also, when it comes to software and applications, limiting choices don’t necessarily mean limited options and features, but it can also mean providing an interface that makes it easy for users to complete their tasks. Google’s search engine is a primary example for this - it has tons of features and growing(oneboxes, define, advanced features, movie, music, etc, etc) but if users had to click a special button everytime they wanted a music search or a news onebox, they would often not use the feature. Google blends it into the main search interface in such a way that sers immediately see its relevance and usefulness; thus, helping users make a good choice.
Imagine the alternative - going to a search engine with 50 buttons for each type of search and search feature. What if you don’t even know where to start; after all, I came to *search*.
Comment by or | May 8, 2006
Nathan,
I was a little uncomfortable with Schwartz’s thesis as well. Sometimes it sounds like he wants to limit choice so users will *feel* better, even if they’re objectively worse off.
He would probably argue that many people truly are worse off with more choices, since choice paralysis prevents them from doing anything at all - like the 401k investor who lets her money languish at 0.25% because it’s too overwhelming to choose among better performing investments.
But Google’s challenge is to figure out how to serve “maximizers” like you and me who want all the options, while simultaneously working well for the “satisficers” who are happy with “good enough” as long as it’s simple.
We can see personalization, clustering, and onebox helping with this, but it’s still far from perfect. How to make search perfect for everyone from the Digg maximizer to the AOL satisficer?
Comment by mahlon | May 8, 2006
Paul Kedrosky had a similar post today arguing against those that want “simplicity” in user interfaces.
http://paul.kedrosky.com/archives/2006/05/08/business_20_sim.html
Comment by Greg Linden | May 8, 2006
I think Schwartz’ argument is more about the cost of information and time spent making decisions than about reducing choice.
Around 24:15 in the talk, he said, “If you know exactly what you are looking for, more choice is better. Because the more options there are, the more likely you will be to find exactly what you want.”
He then goes on to say that if you only have a rough idea what you want, you “try to answer two questions. One, what do I care about in [X]? Two, which one of these things has what I care about?”
Earlier in the talk, he argued that we often are asked to make decisions on “the basis of close to no information.” He is making the point that we often are given a huge number of choices without the information necessary to differentiate between those choices.
How can we deal with this?
One approach might be to show a few good choices first (as Schwartz said, “the importance of defaults”), then give people the ability to explore and learn about all the choices. To some extent, this is what relevance rank already tries to do. The top results are great, but the next 1k results are available if you want to look through them.
Another approach might be to describe all the options available at a high level (as Schwartz said, “structure the options hierarchically”). Some shopping search engines work like this. You rate the importance of selected characteristics, then see what the best match is for those. The focus is on refinement, narrowing your search among the options. Clustering search engines like Clusty might fall into this camp too.
Yet another is to try to learn what you want. Even if you do not know or cannot articulate what you want, perhaps it can be inferred from your behavior, at least enough to target the options more precisely.
Comment by Greg Linden | May 8, 2006
Some choice is ok, but I don’t think you really understand what it would mean to be able to control everything.
Take a fridge, its nice to be able to control the temprature, but do you really want to be able to change the feedback model that controls the motor? Or the rate of coolant flow? Or the amount of air that gets sucked out for the anti-frost thing?
And it’s the same with software, in microsoft word you can drag the file menu to the bottom of the window, or the side, but how many people do you know who have it in anything but the default position?
The vast vast majority don’t care about options at all, and even power users don’t care about them as much as they think they so.
Comment by reednj | May 9, 2006
Are Less Options Better in Search? Barry Schwartz Presents at Google
Nathan Weinberg reports that Googlers Advised To Limit Choice by professor Barry Schwartz, not me - the other one who has his fame from teaching and publishing on psychology, and wrote the very popular Paradox of Choice. Nathan posted his…
Trackback by Search Engine Roundtable | May 9, 2006