UI fun: Better queries

In the last post, we all agreed that search engines have to keep trying different ways to improve search. For example, if you sit down and look at queries that users type, it’s clear that spelling mistakes can be a problem for lots of users. That’s what led us to work on a kick-butt spelling corrector. I love that you can get “Did you mean:” suggestions so that [sturling engine] suggests [stirling engine], but [sturling silver] suggests [sterling silver].

So what else can be improved besides spelling? Well, how about queries? Check out this UI test:

To be or not to be

The user entered the query [to be or not to be] without any quotes. Google is able to suggest that adding quotes to make [“to be or not to be”] is useful. It also shows some sample results (yup, Shakespeare is a much better result). Finally, it even gives you other good searches to try, such as [hamlet] or [to be or not to be that is the question]. That’s a pretty helpful UI test in my book. I’ve also seen suggestions like [mono] -> [mononucleosis] and [24 amendment] -> [24th amendment]. Nice.

Why am I even bothering to write about this? After all, Google is always trying new things to improve search, including different UI tests. Well, an SEO firm based in New York claimed this was “breaking news” and implied that Google was interspersing ads into natural search results. Nope, not at all–this is yet another algorithmic UI test. And the SEO would have known it wasn’t breaking news if they read Gary Price on SEW regularly.

So just for completeness: people who know Google well will go “Cool” and move on. Other folks will ask things like “Are queries selected by hand–can my query get in on this? Is money involved?” And the answer is: it’s all algorithmic. The algorithms pick the queries where this could be helpful. Of course money isn’t involved at all. We’re always running experiments to improve Google–sometimes it’s noticeable, and sometimes it’s not. Don’t even get me started on all the ways we’ve tried using ellipses in our snippets to make them more useful. 🙂

25 Responses to UI fun: Better queries (Leave a comment)

  1. Since you mentioned something related to something I see fairly often I’ll bring it up here. For searches containing numbers you get some strange results for example [top 10 alias episodes]. you see it pulls in things like this:

    Added on: 28-Jan-2004 Downloads: 6562 Rating: 10.0 (10 Votes) … Description: This is the promo for the ALIAS episode that will air on Feb. 23, 2005. …

    French Alias Website : Episode Guide, Rambaldi, Photo Galleries, Goodies. … 10, Alias File French website on Alias ! With photos, news, fanarts, profiles, …

    Ausiello: I stand by my story that Michael Vartan will be back on Alias this season. … “We’ll see,” he added. “Let’s get the mother ship into the Top 10.” …

    “Alias” – Episode 4.5 – Nocturne. Posted by Scott Pepper on February 10, 2005 04:59 PM (See all posts by … On Top 10 Worst Song Lyrics… Ever… by uao …

    I’ve seen it pull in things like ages, dates, measurements, sizes, ratings, and even forum post numbers. Hope thats helpful.

  2. I use the quotation marks very often during my Google searches. It’s to the point where I have them in most of my queries. I think it helps me get to what I need to find very quickly. Some people ask me why I use them all the time and those same people ask me how I find what I’m looking for so quickly. The answer is a combination of both of the last two questions combined with a logical conjunction such as “because”. I love using quotes! But sometimes I wander if my quote habit causes me to overlook something that’s pertinent and possibly useful.

    And about the ellipses…

    I’ve used those in my search as well too. I’ve noticed that most of the time I get the same results if I include “…” after my search words. But there is a noticeable difference if you do a search with the search with the search words and ellipses together as opposed to the search words and ellipses completely inside of the quotes.

    Must be those algorithms. I think it’s ridiculous that someone would think that Google would use spelling errors to determine the placement of ads. But I see where it would be beneficial to someone that would need the grammatical help. Not that my writing is perfect or anything…

  3. Everyone keeps waiting for you to slip – to give into the pressure – to become too greedy.

    And all the while, you keep improving products first and making money second.

    We’ll continue to speculate and form conspiracies because to us, your a freak of nature and we lack comparitive reference points.

    Great stuff here, Matt. ~ becoming required reading for the industry in one week is an accomplishment that only a few in this business can pull off.

  4. Showing additional pages from the same site is a very good idea and will be useful.

  5. There are some speculations that these results are going to be paid ones – Is there any remote chance that Google will include sponsored listings within natural listings ?

    – Sudha

  6. Hi there

    I have a few problems with your post

    First is the reference to SEOs reading Gary Price at SEW as if this should be a daily requirement. Hate to say it but for many of us, what other SEOs have to say, may or may not be of import.

    For myself I wouldn’t sit around and read a certain author everyday, just to see what it is he may have to say. A more varied intake is usually best.

    So it should be noted that a few people may have discovered the new UI and reported it. Cannot expect SEOs to run around site to site and see who has found it first.

    I love SEW but one of the things with having a website so cluttered is important information must be found, as opposed to it being put right in front of the user.

    Are you using on mousedown and these new referal listings to measure if people use then and from that data decide whether to expand its usage?

    Next there has always been the cry of relevancy, so you find it acceptable to offer a user a rappers lyrics when they do a search for “piggy bank”
    to me that seems a bit of a turnaround of prior corp. stance.

    My research on this shows the top 3 positions from a search for “piggy bank lyrics” are then placed in positions 6,7, 8 for search term “piggy bank”.

    Taking this new UI and extending it why not toss in a few referals for “piggy back” ??

    By dong this you have made old positions 9 & 10 obsolete, the traffic from those spots was very little to begin with, wiith this new referal its pretty much game over now for 9 & 10.

    The next issue is webmaster mentality will now go from top 10 to top 5 and for some only top 3 will be good enough so that they will be able to be placed at 6, 7, or 8 for another related term.

    Narrowing the natural SERPS to 5 positions will have more and more people turning to blackhat SEO to game the results.

    Also if this is supposed to benefit the end user so much how come Google is being so quiet about it.???

    Why not just make it a public beta and get the data needed to be collected from the user end….seems it was done this way knowing the seo community would see it first and hhype it up so google can see reaction to it’s implicit commercial & non commercial value.

    On Demand returned Comcast Corp in all three spots and stinks of being a paid placement,.and opposite piggy bank above, doesnt seem to be pulled from another related term 123 as a search for on demand tv, on demand television, on demand services, on demand listings, on demand shows, and others does not return results with Comcast in the top 3 spots..

    This move is totally monetized based, trying to run undercover as a user enhancement (cough cough like Comcast needs Googles help to bolster their reveues)

    Many people looking at Google seem to forget that now as a publicly held company with shareholders they are responsible to, accountants are going to be driving the company forward and not the boys…

    From an accountants perspective the free results listings are a non revenue generating expense ( some will need to think back before PPC to see it this way).as typically there is only 1 2 or 3 pages looked at by most searchers, yet 100s of pages for the keyword query which do not earn revenue.

    Maybe instead of selling those 14million shares, Google should repurchase all outstanding shares and go back to a privately held company no longer under the direction of coporate beancounters.

    Peace

  7. My guess is those Google engineers have been real busy for a while, but since we didn’t actually see the gears turning, we assumed they were just “raking it in”. Things like the sandbox didn’t help the industry perspectives any. Now we see feature after feature coming along – the fruits of that labor?

    It certainly is enough to keep us all busy, provided the search results are good enough to keep the users happy.

  8. “the algorithms did it”

    They’re your algorithms. You wrote them. You run them. The reason Google does or does not do something is a Google decision.

    In this case, Google have decided not to intersperse adds with search results. Why can’t you just say that?

  9. Matt you have got to be kidding. Tacking on a better result at the bottom doesn’t make the top results suck any less! Why not 3 sets of results? 4? 24? All your saying by doing that is that at best you can only deliver half successful results, and leaving them to choose which half!
    Or perhaps you want them to adjust their search technique to your algo, which is even worse.

    Why not do it this way:
    To be or not to be
    to be = verb infinitive
    ‘to be’ appears twice in the phrase, the purpose must be to search for 2 occurances of “to be”.

    i.e. you search for “to be” or not “to be” and only return pages where “to be” appears twice.

  10. Enlightning

  11. Looks like they compiled a list of queries that weren’t answered on the users first try, but did get answered on the second try. Most users would then have refined the query to find what they are looking for. Looks like a simple application for all that statistical data they have compiled!

  12. Clint, I think the point wasn’t that you should only read SEW but that if you are reading a variety of search industry resources — as you’d expect a good SEO company to be doing — you’d have realized from either SEW, Search Engine Roundtable, Threadwatch, WebmasterWorld or other places that this was part of a long-standing UI test. As for being cluttered, the headlines on the blog are front-and-center, easy to spot in my view and if you take the feed, you don’t get any navigational elements, ads or anything.

  13. I have to agree with NotToBeAFanboy, the results for what the user searched should be returned first. At the very least, the ‘Lowercase “or” was ignored…’ should read ‘Google only searched for documents containing the word “not.”‘

    Where I get a bit more confused is that the very first entry is right on the money: “Book results for to be or not to be.” So Google is smart enough to know what the user is actually searching for! Maybe it’s that you just picked a poor example, but I kind of get the feeling that this UI improvement is simply a bandaid.

  14. I just came across a live example of those type of results, by searching on ‘on demand’, and I have 2 comments.

    (1) You need to do work on that ‘additional’ pages from this site bit. Where’s the value to searchers in listing pages like ‘privacy policy’ and ‘register / edit account’ ? Your example was excellent, but Privacy Policy???

    (2) The “See results for …” section is just going to confuse people, imo, and is likely to cause some of them to go elsewhere where the results are plainer.

  15. Why couldn’t you guys just get it right in the first place… when a user types in a query, 9 times out of 10 they are looking for it as a whole.. This is a perfect example why your search engine stinks…

  16. Google does not necessarily show additional results for your query with quotes. For instance if you search for harry potter book it will display results for harry potter book 6 in the same manner. I’ve blogged about this: google’s new toys
    Also, does anyone know why that strange ad for aol appears when you search for book? Do aol use some sort of automatic keyword bidding system?

  17. I can see the logic in offering people corrections to help find more relevant results. Changing to be or not to be into “to be or not to be” seems a valid suggestion. But how is changing ‘on demand’ into ‘Comcast on demand’ a relevant change. Why is it not IBM on demand servers?

    Why do Comcast get special treatment, surely the only possible reason is money. Even if it isn’t a paid advertisement, it sure as hell looks like it and people will think that it is.

    Leaving the reasoning aside, personally, i hate the new UI anyway. Adding a block or extra results into the middle of the page makes it hard to follow and is very distracting. It stops the eye scanning as quickly and if the user meant to search for the query they entered you’ve just placed a whole load of utterly irrelevant results in front of them!

  18. Adam, Comcast isn’t getting special treatment–that’s what the algorithms suggest is the best search for people doing the search [on demand]. And money isn’t involved. You’re welcome to believe that “the only possible reason is money,” but you’d be incorrect. And cynical. 🙂

    Here’s another example: [abc survivor] returns the suggestion [cbs survivor]. (The program Survivior runs on CBS, not ABC.) Did CBS pay us to suggest that search? Of course not. We suggest it because it’s probably more likely to be what the user wanted. But I appreciate your point that the UI could be different and that could make for a better experience. They’ve been playing with different UI’s for quite a while.

    Guy, I wanted to provide the background/context so that people would understand, but you make a fair observation. Just to say it clearly, these are not ads interspersed in our search results (they’re alternate searches which we think are useful) and no money is involved.

  19. Another search which I think shows that google doesn’t tamper with the ranks is “free email” no quotes on google it shows yahoo in number1 position, while gmail is around 23 (which BTW was around 9 before Jagger started) also I notice google pay themselves for ads for that search as well. On yahoo the same search shows Yahoo on number 1 and gmail is around 5?

  20. Sorry that post is in the wrong spot

  21. How come a search for ‘endangered species Africa’ without quotes produces an offer of results for ‘cookie recipes’?

  22. This is why I like Google search. They always try new things to give the searcher the best results.

  23. a very good idea and will be useful

  24. At this url:

    http://www.google.com/support/webmasters/bin/answer.py?answer=47334&topic=8523

    This is what is written:

    “We only show Sitelinks for results when we think they’ll be useful to the user. If the structure of your site doesn’t allow our algorithms to find good Sitelinks, or we don’t think that the Sitelinks for your site are relevant for the user’s query, we won’t show them.”

    How can one determine whether the structure of the site is good enough for the Google to consider it worth as a good sitelink.

  25. I think you do a killer job – considering you take the garbled pile of rubbish most people type in the search box and still manage to figure out what they really mean just about every time is an incredible feat.

    Mark

css.php