Google's Mysterious December Drop Solved? (GOOG)

Peter Kafka | January 24, 2008 6:50 PM

google_logo.gifWe're still curious about two reports we've seen in the last week -- one from Nielsen, the other from comScore -- that show Google losing a tiny piece of the U.S. search market in December.

They are small drops -- Nielsen shows a 1.4% decline, comScore shows a 0.2% loss -- and we don't want to overplay them. But shouldn't they mean something? Doesn't Google's market share always get bigger?

Not necessarily. A person who follows search stats more closely than we do argues that Google's December drop is probably easy to explain: Students spend less time on their computers in December because of the holidays.

Hmmm. Shouldn't that phenomenon affect all the engines equally? No, says our search guru -- Google's share of students is much higher than its older rivals, so it gets hurt a bit more. And the same thing happens every spring, when school lets out for the summer: Last year comScore showed Google losing 1.6% from May to June. Then its share began climbing once again. So: Do you find this convincing? Or do you think something else is at work here? Let us know in comments.

comscore-2007-search-share.gif

Subscribe to SAI: RSS, email newsletter (no spam)

More:

Share:
Post to Del.icio.us


< Prev. Story
Next Story >

11 Comments

anon said:
how about, people are so busy celebrating the holidays that the last thing many do is actually sit at a computer?
Danny Sullivan said:
A slight drop from one month to the next pretty much means nothing. If you don't see it consistently for several months in a row, it's indeed worrying for no good reason. In addition, you also have to consider the actual number of queries as well as share. For example, Yahoo has had months where the share is down but the number of searches is up, because the overall pie expanded (say Microsoft runs a content and generates brand new searches, for instance).

More caveats like this, as well as trends over the last six months from four ratings services here:

http://searchengineland.com/071228-173523.php
Stephen Fraser said:
ALL web sites see a significant drop off in traffic at the end of December. E-commerce sites get a boost because early December numbers are higher than usual. Google's volume of searches is so high that the phenomenon is more noticeable.
FM said:
This is beginning of the end for Google. Honey moon is over. While Im not saying theyre going to loose their market share entirely, it will certaintly come down. In the near future, new technology will come out to do to Google, what Google did to Yahoo. Sell Goog while still in 500s.
wiseGEEK said:
microsoft had a huge advertising campaign during the q4 driving lots of traffic to its search property. vibrant media (intellitxt), for example, had a huge contract with them that expired on 1 jan 2008. my understanding is that microsoft will resume their aggressive advertising in q2 - so it is reasonable to assume that their share will drop in q1 but rebound in q2
jim said:
CV,

You are an idiot. Even the smallest panel in the world wouldn't misread a site so large and the trend on a metric with such high incidence. A panel of 2MM people that is balanced for demo/geographic factors would measure with high precision a simple measure such as searches.

Please try and think before responding.

Dumb ass.
Dorian said:
Plausible, but seems incomplete -- too facile.
CV said:
So many factors that can drive search volumes. Maybe it was college students on break, maybe it's a combination of slow news days, or the lack of celebrity mishaps. The drop itself is not significant. We also need to remember that both reports come from panels, meaning that how much of a drop off there was (if there was one) came from the panels. Before one starts crying that end is near for Google, a closer look at each panel would need to be in order to make sure there were no anomilies with their data. The only way to truly know if Google saw a drop off is for Google to release their own internal results. Then we could compare those numbers to last month as well as last year and see if there really is cause for concern.
Paul said:
Makes sense to me.
http://nextthing.wordpress.com/

Join the discussion



Copyright © 2008 Silicon Alley Media, Inc. All rights reserved. Registration on or use of this site constitutes acceptance of our Terms of Service and Privacy Policy.