Back when I was reporting the book, I remember a meeting I had with Gary Flake, then the lead technologist at Overture, now a Fellow at Microsoft running Live Labs, responsible for stuff like Seadragon, Photosynth, and now, Pivot, an experimental approach to large datasets that attempts to rethink some fundamental approaches to what we understand search to be today.
Back in 2004, I asked him why we couldn’t move forward in search interface, which struck me as a major issue (and still does). Gary looked at me ruefully and said something I’ve never forgotten: “If only I had just one modal dialog box…”
What he meant was that search, at that point, was a race for the best ten blue links, and anything that got in the way of that, like a modal dialog box that popped up and asked a refining question, would mean that a very large percentage of folks would abandon the search.
And abandonment of the search meant loss of revenue.
Google was just better at getting (approximately) the right first set of blue links, and hence, it won the first round of the search interface wars.
But things are changing. A lot.
I’ve notice lately that I’ve not been happy with search results, because, well, there’s just not enough refinement in the SERPS. But it’s not just the SERPS, its also the interface. I’ve written a lot about this, but in short, I’m frustrated with the way search does post declarative navigation. (OK, that’s totally geeky, but those of you who really care probably know what I mean).
And this is why I’m grokking Pivot right now, and let me just say this….this is worth grokking. So I am going to be doing just that over the next day or so…expect more soon.
If you’d like to grok Pivot, check out this presentation. You’ll need Silverlight….
You had me until you said “Silverlight”… moving on
You can see a 4 minute long introductory video at
(which does not require silverlight).
Pivot is way cooler than the bloated Google Wave.
All Microsoft has to do now is find out how to integrate this amazing App within the Bing Search Experience.
What he meant was that search, at that point, was a race for the best ten blue links, and anything that got in the way of that, like a modal dialog box that popped up and asked a refining question, would mean that a very large percentage of folks would abandon the search. And abandonment of the search meant loss of revenue.
The problem here is that there is a tradeoff between refinement and ads. More time and effort and attention payed to refinement means less time and effort and attention paid to ads.
I would venture that abandonment is a red herring, because better refinement might lead to the exact same outcome: Users paying more attention to refinement means users are clicking fewer ads — a loss of revenue even when there is no abandonment.
I’ve been making this point for years now — that ads are prioritized over refinement — and I’ve never seen a compelling refutation. (Here is one old comment; I could easily dig up half a dozen more going back years, if anyone cared: http://battellemedia.com/archives/004316.php#comment_129348 )
Consumers are finally starting to demand refinement, and so the search engines are slowly being forced to give in with tiny steps. But at the end of the day, the more time the user spends in the SERPs (and that includes refinement) the less money that the search engine is going to make. That is just a basic fact, given the current search engine business model. Abandonment or not, the business model itself does not like refinement.
Going forward, that will continue to be a problem.
The UI is nice, and faceted search for the big, messy web has always seemed like a very cool – if distant – promise. But I’m having trouble connecting Flake’s examples to the kinds of things I actually search for. Very rarely do my searches have anything to do with perusing a large data set or collection of similar items.
Most of the time, I use Google to get to one thing – often a site for which I can’t remember the URL, or simply don’t feel like typing it into my address bar. For these, Google does a great job of not only putting my target website at the top of my search results but often displaying a list of deep links to the important parts of the site.
Second most often, I’m looking for the best example or implementation of something for which there are only a few. Recently I tried to find an online version of the classic video game “Lemmings” and a video of the opening theme of “The Greatest American Hero.”
Third most often, I’m looking to educate myself about something. In these cases, I’ll scan a few articles, documents and blog posts that come up in the results and bookmark a few for later reading.
The examples from my own habits I can imagine using something like Pivot for are digital media (pictures, videos and audio files) searches, shopping and travel planning. For these cases, Pivot might be a bit of an improvement on current user experiences (offered by the likes of Google Advanced Image Search, Kayak, and lots of e-commerce sites), but it doesn’t seem revolutionary.
I am glad to find this blog, Keep up the good work.
Cheers.
–Google Money Kit
JG makes a strong point, and a worrying one that the search engine business model is opposed to refinement. I guess this’ll open up the market if someone can come up with a way of generating revenue and refinement.
Gary Flake is wonderful. I like to listen to him and I think he will develop himself in a correct direction.
how to convert m4a to mp3