Tuesday, November 27, 2007

Google snippets

Matt Cutts visited us here in Seattle by heading to Google Kirkland office. They had decided to make a few videos of Matt while he was there and post it to the Google Webmaster Blog - Anatomy of a Search Result. Matt's video explains the snippets (title and description that shows up in your search results). He used Starbucks as an example. The Starbucks site is a good one since it has limited content on the page, but uses meta descriptions in their code.

Matt mentioned in the video that we have no control over the site links presented underneath the snippet which isn't entirely true. While we are unable to pay for the extra sitelinks within the snippet, we are able to control them through the webmaster tools. Under the "links" and then "sitelinks" within webmaster tool, if Google has generated sitelinks for your website, you are able to control them from there.

Always remember how the snippet works:

  1. If the search term is within the meta description of the web page then the description will appear in the snippet (with the term highlighted in bold).

  2. If the term isn't in the meta description or there isn't a meta description then the words around the content within the body of the page will be displayed (with the term highlighted in bold).

  3. If the page does not have content and does not have a meta description, then the description is generally pulled from the open directory project

  4. If there isn't any content on the page, no description in the meta tags, and the site hasn't been submitted to the open directory project then the descriptiontion will be blank.

A great example of the fourth scenario is a client I have been working on that came to me with his website that was created entirely in graphics. The website was a great design, and had a lot of great information on it, but the text was put in images and placed on the site that way. The website wasn't even ranking for the name. The only way to pull it up int he results was to do a site: search to see that the pages were getting indexed, but the content wasn't getting recognized (because of the images).

I am currently in the process of pulling out the text from the images and the html is recoded to keep the design the same while keeping the text still recognizable. I also added a webmap for more efficient indexing, and landing pages for the terms that the client wanted to rank for. There is an xml sitemap submitted to the webmaster tools, and a Google analytics account setup for tracking conversions.
The client should start to see better results soon after we get the new site launched.