And the Interface Evolves

…toward conversational interfaces….news from Google acknowledging that ten blue links is getting old (we knew that, so did Google, but…): There have been a lot of recent improvements to web search, but the appearance of results themselves has been pretty constant — 10 or so web pages in a…

…toward conversational interfaces….news from Google acknowledging that ten blue links is getting old (we knew that, so did Google, but…):

There have been a lot of recent improvements to web search, but the appearance of results themselves has been pretty constant — 10 or so web pages in a vertical list. Frequently this is exactly the right format, but for some searches you need more options and more control. That’s why we’ve created our experimental search page to let you try out some of our newest ideas.

You may have noticed our “alternative views” experiment showcased last May. This lets you visualize your search results in new ways, and we’d like to highlight some of the features we’ve recently added.

8 thoughts on “And the Interface Evolves”

  1. Let me give the guys a couple tips:

    1. clustering:

    see e.g.

    2. menu-driven (aka “suggest”) — KISS:


    3. is it LIVE or is it MEMOREX?

    … … … hmm… — let’s see… maybe:

    cf. , , , , … ?

    4. if you want to refinance, try — if you’re not happy with the results, then maybe try or or …

  2. LOL @ ooooh… ahhhhhhh!!

    Some are pretty cool like the Google Image Labeler, which can show you how incredibly dumb or intelligent you are within two minutes (how did Het is al van u score 20000010 points?!!), but I others confined to 1024×768 screens would probably find them a bit too cluttered.

  3. we knew that, so did Google, but…

    Are these just the same four or five different visualizations that some other commenter a few weeks ago recommended that I take a look at? Yeah, I think that’s the page.

    Good for Google for throwing these things against the wall. I laud that. But this is hardly a commitment to an evolved interface. I want to see these things, and more, live. Not on some hidden, experimental, labs page that the majority of searchers aren’t going to find, much less even know to look for.

    What I find sorta odd about the experimental page is that Google says “you can only join ONE experiment at a time”. That doesn’t make any sense. If you want to move beyond 10 links, you gotta realize that different queries necessitate different visualizations.

    For example, if I type “iraq war”, then I might want the timeline to automatically appear so I can see events on the timeline. On the other hand, if I type “vacuum cleaners”, then I might want the automatic query expansion / right hand contextual search navigation to appear so I can see various types of vacuum cleaners, manufacturers, reviews, etc. A timeline would be of little use to me. Finally, if I type “tim burton movies”, then I might want to see both query expansion (all his movies, plus related persons such as Johnny Depp and Danny Elfman) as well as a timeline view (when his movies were released).

    So for Google to only let one of these interfaces be active at a time doesn’t really make a whole lot of sense. They’re basically requiring me, before doing every single query that I do, to go in and turn off one interface, then turn on the other, if my query is of a different type than the interface I currently have active.


    Let’s see these things start to appear on the main Google page, and in an automated, helpful manner, i.e. when it is contextually appropriate to my search activity. Don’t force me to go activate/deactivate it before and after every search.

    I want something better. And I want it before the next ten years pass by.

  4. Also: I was just thinking that the map user interface might be especially useful if users could select from a drop-down list of sexual positions, but without some “authority control” I fear that it will be only of limited use….

    🙂 nmw

  5. Also (again, this is “also, part 2” 😉 :

    The more Google fiddles and fuddles with their presentation of the results (and thereby again continues to drift away from the original impetus that — way back when [originally] — made Google great), the less the click stream reveals about what’s “out there” on the Internet (and that holds whether the what’s “out there” is the so-called user [“recipient”] or whether it’s the so-called provider [“sender”]).

    Now that I think about it, I guess I am echoing (and emphasizing) JG’s comment above: The criteria Google uses to rank results is saying more & more about Google’s “preferences” — i.e., what “matters” to Google (and, as JG has ably argued: what Google thinks might matter to users/providers ain’t necessarily so).

    This is the basic pitfall of a “one-size-fits-all” search engine (and it’s also why “educated” / “advanced” searchers scoff at such a “hammer & nail” approach).

    Here’s an example: if you want to rent a video, it used to be that you had to know whether it was in VHS or Betamax format; and/or if the data were formatted according to PAL or NTSC — and now you have to decide between Blue-Ray or HD DVDs. However: None of this matters if you want to buy a bicycle.

    And here’s a little “BTW” for Mr. Donahoe: If you are looking to make a one-size fits all search engine, you too will send your company into oblivion — and it is better to recognize that simple fact sooner rather than later.

  6. Well, I feel the experimental interfaces for Google results are not that creative and well thought of. For example, it would be foolish to have a map interface for someone looking for information on new trends in genetics.

  7. Nice experiments. But IMHO the 3 columns-approach of (my current fav) and the menu-driven Yahoo! are more useful to me. Especially Ask, how good the auto suggested content is. By far better than the Google experiments.
    But maybe this is also kind of a personal descision whether a displayed information will become an additional value or just a waste of space.

Leave a Reply

Your email address will not be published. Required fields are marked *