After our presentation for the Magazine Publishers of America in NYC on Tuesday, we received some requests for additional research.  I spent this morning trying to find some statistical explanations for the performance of the magazine websites by running some regressions, checking for correlation between the presence of Web features and traffic (as measured by Alexa).  However, much like our regression analysis of newspapers , this did not result in any significant results.

I wanted to find out which websites were performing the best in relation to print versions of magazines, so I created an online-print differential.  This figure represents the difference between the ordinal online ranking of a magazine website and the magazine's ordinal print ranking.  Large negative numbers indicate a website that outperforms the print magazine, where positive numbers indicate print versions that are outperforming their online counterparts.  Here is a table that shows our results.

As you can see, it's difficult to infer any patterns from this data.  I did see a few interesting things, however.

• The Popular Science website has the highest online-print differential of all the magazines we researched.  This makes perfect sense, considering that the demographic to which this magazine appeals is most likely tech-minded, therefore more likely to use the Web to access PopSci materials. Also, the Popular Science website is relatively full-featured and robust.
• Women's magazines made up the majority of the magazines at the bottom of the differential list, meaning that the print versions of these magazines drastically outperform the websites.  I guess I'm not really sure what to make of this, but I would be interested to see if there is any pattern here if we compared this stat to general usage rates for women vs. me
  n.
• I was surprised that TV Guide had the differential it did.  TV Guide has a great website, so I found it odd that its website didn't fare better in this analysis.  I guess this can be attributed to the fact that TV Guide has a massive print audience.
• Endless Vacation is another surprising case.  This website is primarily a brochure site with no online functionality.  Yet, the print version of this magazine only slightly outperformed the website.  What exactly are people looking at on this website?  I can't figure that one out.

Overall, this analysis did not present any clear conclusions.  I think that a problem with this type of research is that obtaining data for magazine Web traffic is difficult.  I used Alexa to determine the traffic rates for magazine websites, but I am not confident in the true accuracy of these figures.  Since Alexa only tracks websites visited by people who have installed the Alexa toolbar, the results are probably skewed toward tech-minded audiences and are not an impartial measurement. 

Another problem with this sort of analysis is caused by what we dubbed "integrated sites"–magazine sites that are part of a bigger network of sites.  Redbook , Country Living , Money , ESPN , and Sports Illustrated are examples of this type of site (Redbook and Country living are part of the iVillage network; Money and SI are part of CNN , and ESPN the magazine is combined with ESPN the network).  Alexa could not read these magazine sites independently, and only presented data for the umbrella network.  For example, when you type the Redbook URL into the Alexa search function, it gives back data for the entire iVillage network rather than traffic just for Redbook.  As a result, we left these "integrated sites" out of the analysis, which probably left holes in the data.

While the analysis itself may not reveal any earth-shattering conclusions, it does draw attention to the lack of transparency in the industry.  In order to conduct a real analysis, we would need access to accurate traffic and page view data, which would probably have to be acquired from the magazines themselves.

Check out our magazine study here and our newspaper study here .