How Does Changing Website Content Improve Search Engine Page Rank?

The NewsBox - Relevant News Stories

Our company’s product, the NewsBox, was used to supply the test website with changing content. The NewsBox can be sized otherwise customized to fit in an empty space in most websites.

A human editor selects stories based on the number of keywords in the title. Most of the stories had two keywords in the title. The editor also checks that the story should be interesting and relevant. News stories are located through news feeds and manual searching.

A PHP Include statement places the story’s date, title, publication, and link in the HTML code thus exposing this content to search engines. A sample of the embedded content is shown below:

The website designer selects the size, shape and formats of the text. The text automatically scrolls at a rate to eliminate browser “jitter” resulting in a smooth display.

The scroll bar only appears when the user’s mouse hovers. Stories display as shown below in the sample NewsBox.

Each Test Website Has Fictional Products

Two non-existent products were described using clip art photos and non-meaningful text.. One was an “Undersea Radio Galvanometer” and the other an “Aerial Chronograph Observatory.” By using these generally non-related keywords, we hoped to keep organic searches to a minimum1 since monitoring key word searches was not part of the test.

By eliminating all other variables, we wanted to how the site appeared to Google and how they would react to the one difference between them.

Devising the Test

To demonstrate the effects, we built two websites that were structurally identical with only the addition of a News Box to one of the sites.

For this test we were only interested in how Google’s crawl behaved differently. Nothing would be changed on the two sites except the News Box.

Using clip art photos and technical specification style text two nonexistent products were described.

Site A: www.ChangingContent.com
Site B: www.UnChangingContent.com
Pages per site: one
Word count: equal, 220 words
Graphics: equal, three
Backlinks: four, one to each site. Each linked site posted both addresses to equalize weight of back links
Domain life paid for: equal, two years
NewsBox Updating: fifteen new stories, every two weeks.
HTML/CSS: same except for code snippet for NewsBox in one site.

During the test no site content was changed except the NewsBox. No other variables such as back links, amount of content, or any other variable was changed. The keywords were selected to be non-competitive with other sites.

If you keyed in “radio undersea galvanometer” the site was #1 from the first week. This was good since we weren’t measuring keyword ranking and didn’t want many natural searches for the fictitious products. We weren’t looking for natural searches in this test.