« Speedwagon Nominated for Best of Search Blogs Awards | Main | Who's Doing the Work Here, SEM Friend? »

Watching MSN's Algorithm at Work

December 13, 2005

I was all ready to bash MSN's search algorithm a week ago, and I'm glad I waited. Instead, I've grown more impressed with its self-governance and, I daresay, its audacity.

During a routine vanity search for [seo speedwagon] about two weeks ago, I was shocked at the top result - an actual results page from Google Blog Search. The following screen shot was snapped on 12/4 and shows the MSN results page for the query [seo speedwagon].

As you can see, the top result is a dynamic results page from Google's blog search engine. And that's my fault, I guess. Back in September I wrote about Google's new blog engine. In that post, I talked about (and linked to) various test searches I had run, including a search for our blog name. No big deal.

When you think about a search results page like this, it has several factors that, in a vacuum, should make it rank well:

• A succinct title with relevant keywords
• Loads of on-page instances of keywords
• At least 20 outbound links to related sites (Google Blog Search results pages link to both the specific post and to the blog's home page for each of the 10 positions.
• (In this case) four incoming links (albeit from the same domain) with targeted anchor text. Because of the way the Speedwagon is written and archived through Movable Type, each post tends to show up multiple times. MSN sees this specific post in four locations:
  • The post's individual archive
  • Our archive of blogging articles
  • Our monthly archives for September
  • Our archive of Google articles

So ironically, I created the very page that supplanted our own site in a vanity search. How poetic.

Now here it gets even more interesting. We can debate the merits of the Google Blog Search results page all day, but in truth, MSN never should have crawled and indexed it in the first place, because Google Blog Search results pages are specifically disallowed in the robots.txt file for that subdomain:

So let's sum up:

• MSN crawled and indexed a page from Google's blog search engine against Google's robots.txt exclusion.
• Its algorithm deemed that page worthy of ranking for a relatively uncompetitive term.
• A week after I noticed MSN ranking the Google page at position 1, it had dropped to position 2. Today, it doesn't rank anywhere.
• The Google Blog Search results page has dropped out of the MSN index.

I'd say that just about balances out right.

All posts by Erik Dafforn
posted by Erik Dafforn at December 13, 2005 04:00 PM
Intrapromote: [ Case studies | SEO services | Bios ]

Printer-friendly version

Trackback Pings

To TrackBack this entry, use the following URL:
http://seoblog.intrapromote.com/mt-tb.cgi/141

Comments

Posted by: buyers web at April 20, 2008 02:27 AM

Post a comment

Name:

Email Address:

URL:
Remember Me? YesNo

type the word 'dog' below, without quotes (required):

Comments: (you may use HTML tags for style)