Rabid Wombat |

In case you’re wondering, yes, I did actually need this information at the time of query.

I read something earlier this morning that reminded me of a topic that I had wanted to bring up earlier: the speed of search results. It’s an old post, but still very relevant. Basically, Google ran an experiment that increased the number of search results per page from 10 to 30– but the half-second slowdown that accompanied the change caused a 20% drop in traffic among that segment. TWENTY PERCENT!

It reminded me of something that I’ve been meaning to mention on here: Live search is quite a bit slower than Google. Maybe I only notice it because I have a habit of searching both simultaneously, but for me it’s noticeable nonetheless.

To the left is a screenshot I took of a really random query to both Live Search and Google. As you can see, Live Search, which should be showing on the left side, has not yet loaded, but the Google results are already displayed entirely.

To be fair, I hit “stop” in my browser in order to have time to take the screenshot– but that doesn’t change the fact that Live is significantly slower.

It may not be scientific, but I can tell you that I have a tendency to look at the Google results first– and now maybe I have my explanation as to why. If Google shows up first, how can I look at anything else?

I’m not sure if this affects anybody else who does not search both engines simultaneously, but if Google’s results are any indicator, then Live’s split-second delay may be costing them more than they think.

By the way, did you notice that Google’s logo looks a little bit smaller than usual on the search results page? Compare the above screenshot with this one from a week ago. The old one is a little bit larger. Could this be another optimization to speed up the search results? Seems likely.

Apparently Live Search does have a chance! At least… if Google dies. So Microsoft just has to kill Google to take market share from Google… It’s a bit of a catch 22, but I’m sure they can pull it off.

Today I decided that since I never seem to have time to make it to the Apple store, I might just buy a case for my iPhone online. (Apparently the case for an iPhone 2G is incompatible, but that’s another story.)

So with that goal in mind, I searched for “iphone 3g case leather clip” (without quotes). I was hoping to find a leather iPhone 3G case with a belt clip. That’s pretty much what I found, but Live also decided to give me links to “Top Videos” related to my query. See below:

This makes me wonder if Live is actually just taking the word “clip” and deciding that all queries with that word will be linked to videos… But why wonder when you can test?

Luckily, when I searched for “belt clip”, Live did not think I meant anything video related. (Oddly, when I searched for “snl clip”, Live did return videos, and Google did not.)

In the end, both engines offered a bunch of good links, and related ads. Both of them served my needs, and for that, this one gets chalked up as a draw.

Still, Google seemed to know that I was in shopping mode, and Live did not.

If you’re wondering how the iPhone case story ends, I have not made a purchase as of yet. All of them I’ve seen looked pretty worthless.

I’m a Perl coder. Sort of. I’m still learning. That’s why I search for syntax examples every so often. I’m looking for an explanation of how things worked. Well, recently I searched for “map $_ => $_” (without the quotes), some Perl syntax I’d seen written somewhere else. I was hoping to find some information about the “map” function, and how it would interpret that particular set of input.

The results were very disappointing. Live Search returned a map of Redmond (where I conducted my search), and offered nothing related to Perl syntax in its results. I shouldn’t be surprised, this has been a problem before.

Here’s what I got:

It looks like Live is completely ignoring special characters, and is using the presence of “map” in the query as its exclusive trigger to show instant answers. But of course, I didn’t search for “map” at all…

Google returned some Perl related results, although they weren’t entirely useful. Still, at least there my actual query hit the server.

While studying for my Psychology class, the textbook made a claim that most people could not remember all the details on a penny, even though they’d seen them thousands of times in their life. Of course, I disagreed, and did what anybody would do. I sketched what I could remember, and then I did a search.

I decided to use the main search because it’s more accessible– I don’t have an image search plugin for my Firefox search bar. So I entered “penny” and I was off.

I was pleasantly surprised to find that Google was already sure that I was there to see pictures of pennies. I’m not sure how they knew, but they were right. Here’s what I saw:

So why doesn’t Live do this? Is it just the lack of universal search functionality that Google has implemented? Or do they also lack the data and/or logic to determine that many people searching for “penny” may be interested in seeing pictures?

* Disclaimer: I say that many people searching for “penny” are probably interested in seeing pictures, but I don’t know for sure. Maybe Google is wrong, and I’m also a fluke…

In the course of my day I was asked a question about the right syntax for Adcenter to insert the Match Type into the destination URL.

Somebody asked me if this looks right:

http://www.domain.com/trackingurl?kw={keyword}&match={MatchType}

I answered that it looked right… but I wanted to make sure, so I searched for {matchtype} adcenter insert. Here’s what I got:

I was fully expecting to see examples of this syntax right within the snippets of the search results, giving me the only confirmation I needed. Alas, Live had not a single reference to the bracketed word, while Google had numerous examples.

…until now.

I thought it would be fair to highlight that if you were to compare Live to Google every time you search, as I do, then you would find that they often tie, rather than one set of results being clearly superior.

Example query below:

I would wager that this kind of tie is fairly common for many queries. Unfortunately for Microsoft, a tie in relevance does not convert Google users, nor increase loyalty ofcurrent users.

This one’s a draw.

So the other day I had a song stuck in my head. As it turned out, it was a song that I liked. The trouble was, I didn’t know the name of the song, or the artist that wrote it. It had been months since I’d heard it, but I wanted to hear it again. Luckily I had just upgraded to the iPhone 3G, so I could now download the song from iTunes even though I had no WiFi available.

Except I didn’t know the name of the artist, or the song title. BUT, the words of the chorus kept echoing in my ears: “Come back to me, it’s almost easy.” So like any good surfer, I tried searching for those lyrics in order to determine the artist name, and song title.

Using my trusty Firefox toolbar, I performed a search for: come back to me it’s almost easy. Here’s what I got:

Blam! Thanks to Google, now I know the both the artist name and the song title. I was sorely disappointed to find that Live returned nothing related.

But this got me curious… surely with such a large index they’ve got to have the same page as Google does. So I decided to search again for the same thing, only with quote marks: “come back to me it’s almost easy”. Here’s what I got:

Much to my surprise (and chagrin), Live indeed had the page indexed, they just failed to rank it properly. If it was outside the index, at least there would have been a good reason for not showing it the first time, but alas, it should’ve been there.

It looks to me like Live is not defaulting to phrase match, which I see as a major problem.