Rand posted a fun article on comparing website analytics with competitive intelligence metrics. Being in the search and analytics fields, I found his data very interesting. However, I’d argue that these sites are not ideal for comparing the two. Non-SEO focused sites might have higher correlations between the two sets of data because SEO folks are more likely to play tricks with links, Alexa, Technorati, etc.
Regardless, I thought I’d take Rand up on his offer to post new information given the data he shared. The first thing I did was to hone in on the sites that get at least an average of 100,000 visits. I wanted to eliminate the noise of sites with 1-79k pageviews per month, which might be skewing the data. This leaves me with 8 higher traffic sites to measure upon. The correlation results I get using the same method Rand did results in this:
- Alexa Pageviews (0.70)
- Technorati Rank (0.53)
- Alexa Rank (0.49)
- Ranking.com Rank (0.32)
- Yahoo Links to the Domain (0.23)
- Bloglines Subscriptions (0.121)
- Number of Technorati Links (0.120)
- Yahoo Links to the Blog URL (0.119)
- SEOmoz Page Strength (0.04)
The rest had negative correlations, with NewsGator Subscribers being the worst indicator among the bunch. By focusing on the higher traffic sites, we see different correlation scores.
Then I thought to myself, “there must be a better measure.” So I thought long and hard and came up with the following formula: 100-[character count]
The character count is the # of characters in the URL for each site, meaning sites with less characters to type in will receive a higher score. SEOmoz, with 22 total characters in its full url string, would receive a score of 78, while SEJournal would only receive a score of 65 (35 characters in the full URL).
Matching up my new character-count competitive intelligence measurement, I receive a correlation score of 0.90, beating out all the other measures presented!
Advice to any SEO hoping to obtain traffic to their blog: simply go with a shorter URL.