Interesting that Google has let the Times behind the curtain to interview Amit Singhal and Udi Manber, two of the high priests of Google’s most important work – search ranking and quality. Saul Hansell, the author, called me for some thoughts and was kind enough to include a few of my thoughts in the piece.
Google recently allowed a reporter from The New York Times to spend a day with Mr. Singhal and others in the search-quality team, observing some internal meetings and talking to several top engineers. There were many questions that Google wouldn’t answer. But the engineers still explained more than they ever have before in the news media about how their search system works.
Indeed, for those of you who geek out on this stuff, there are clues to a number of things in here. Witness:
In 2005, Bill Brougher, a Google product manager, complained that typing the phrase “teak patio Palo Alto” didn’t return a local store called the Teak Patio.
So Mr. Singhal fired up one of Google’s prized and closely guarded internal programs, called Debug, which shows how its computers evaluate each query and each Web page. He discovered that Theteakpatio.com did not show up because Google’s formulas were not giving enough importance to links from other sites about Palo Alto.
It was also a clue to a bigger problem. Finding local businesses is important to users, but Google often has to rely on only a handful of sites for clues about which businesses are best. Within two months of Mr. Brougher’s complaint, Mr. Singhal’s group had written a new mathematical formula to handle queries for hometown shops.
There is also an interesting discussion of freshness and how that relates to ranking (watch for mentions of “QDF”). Worth the read.
Nice clincher quote. Love that “secret sauce” approach — was that McDonald’s or Burger King?
Seems like the solution is overkill. Why change your whole algorithm, when the query itself is best served on a different type of search interface? On local.live.com, I can simply type “teak patio [tab] palo alto” and get the correct result ranked at the top. No need to muck with an algorithm, when using the right tool (maps-based search) would have been much easier.
But maybe that’s the whole point. Maybe that’s the whole vision behind universal search. Instead of requiring the user to use the tool created specifically for the purpose of locating a business in a city (e.g. local, maps-based search), Google thinks you should just be able to type it into the general search box and still come up with the answer.
It is a decent vision, I suppose. But I think that teaching the user to know when to use the right tools is, in the long run, much more beneficial than always having the system guess what is really going on. What’s that hackneyed saying? Give someone an automatic search result, and their information need will be satisfied for a day. Teach someone how to search, how to use the right tools at the right times in the right ways, and his information needs will be satisfied for a lifetime.
I know I’m going completely against the grain here, but it sometimes really puts me off that removing all thought effort from the user is the approach that is being optimized for. It’s like information fast food. It’s easy, it’s quick, and it tastes good. But in the long run it has almost no nutritional content, i.e. it does not encourage exploration and diversity. And this is how we are teaching the average information consumer to feed him/herself? Harumph.
JG,
Did I read your post right? The right way to solve for this problem is to “teach the user” to do more work? We should pay more attention to critical thinking, not which search tool we need to use for different types of searches. The tool serves its user, not the other way round.
Did I read your post right? The right way to solve for this problem is to “teach the user” to do more work?
No, you did not read it right. I am saying that the right way to solve this problem is to teach the user to work smarter, not harder or more.
We should pay more attention to critical thinking, not which search tool we need to use for different types of searches.
I agree with you; we should absolutely pay attention to critical thinking. But right now, the Google approach is encouraging the user to work dumber, to be lazy and uncritical when issuing a query. As a result, I think the user is, in the long run, going to have to do more work. Not less.
For example, if the user has become so used to typing “palo alto teak patio” into the general search box, in order to get a store named teak patio in palo alto, and “walla walla car parts” to get auto supplies in Washington state, then when the user types “new york dolls” to buy his child a present while on a business trip to the City, he is going to be very surprised when the results come back with punk banks and strip clubs (as they currently do). At that point, seeing that the query failed, the user is going to have to go back and type a second query, “dolls in new york”, and will in the end have done more total work than the user who was smart enough to simply click the “maps” tab at the beginning of the query, and then type “New York [tab] dolls”.
The tool serves its user, not the other way round.
Exactly. And if the user does not know how to use the tool, or does not know which tool to use, that tool is not going to do a very good job of serving the user. One can always hammer a nail into a board using the butt end of a screwdriver. But that doesn’t mean the screwdriver is the best tool to select for the job. The user that doesn’t know this is just going to reach for the screwdriver every single time, whether that user is trying to hammer or saw or drill or whatever. Search engines, whether it’s Google or anybody else, should not be encouraging this sort of behavior. It ultimately places more burden on the user.
I do understand, however, that I am in the extremely small minority with this opinion.
Here is one more analogy: Google is trying to be the swiss army knife of all information access, organizing every single type of information from a single, one line text box. Can a swiss army knife do (pretty much) everything? Let’s say yes. Can it do it all well? Not really.
Think about trying to change the engine on your car. A swiss army knife has all the basic components that you need to change the engine. But by the time you actually succeed in changing the engine using the knife, you are going be extremely frustrated. It will be a helluva lot of work. Rather, it would be much better if you knew that you should be using tools dedicated to the task of changing your engine. A tool that is dedicated to that task will serve the user much better than a general one, and make the whole process much easier, much less work, for the user.
Extend that analogy to search. That’s really all that I am saying here.