Your most popular page is likely NOT to be your most popular page

Summary:
By defining a “page” as a portion of content - and thus defining “most popular” page as a ranked list of consumed content, it is very likely that the most popular page (as listed in your Web analytics tool) is NOT the content that is consumed the most, due to Social Media content spread and consumption.

This post is written with the Online Business Measurement Quadrant in mind. A model describing that controlled on-site measurement is simply not good enough if one wants to get a complete picture of real consumer content cunsumption, engagement, sentiment and so forth.

I will be using this analytics blog as an example to test my hypothesis - that my most popular page is very likely NOT my most popular page.

Example site and content:
Site: http://visualrevenue.com/blog

Where the immediate conclusion from my Web Analytics tool is the following:



Which of course is very much in line with what I reported for the 18 most popular web analytics post of 2007 post. Putting that into the context of the Online Business Measurement Quadrant we get the following subset of results:


Controlled On site Content
Traditional marketing messages published on the corporate website, micro websites etc.

TOP 4

• 879 views - http://visualrevenue.com/blog/2007/07/what-to-measure-on-social-networking.html
• 789 views - http://visualrevenue.com/blog/2007/03/long-tail-and-how-to-calculate-missing.html
• 587 views - http://visualrevenue.com/blog/2007/07/tracking-rss-subscribers-via-img-tag.html
• 437 views - http://visualrevenue.com/blog/2007/06/web-reporting-vs-web-analysis.html

Controlled Off site Content
Traditional marketing messages published on third party website sites, widgets etc.

TOP 4

• 1584 views - http://www.slideshare.net/dennis.mortensen/web-reporting-vs-web-analysis/
• 1025 views - http://www.slideshare.net/dennis.mortensen/indentify-increased-organic-search-traffic-opportunities-in-a-drooping-tail/
• 375 views - http://www.slideshare.net/dennis.mortensen/regional-web-analytics-level-of-advancement
• 110 views - http://indextools.com/docs/IndexTools-Reporting-vs-Analytics.pdf

Uncontrolled On site Content
Brand commentary, Ratings and other User Generated Content published on the corporate website, micro websites etc.

• N/A*

*The only User Generated Content I have implemented is “Comments” – and these are counted as part of the post view and not separately.


Uncontrolled Off site Content
Brand commentary, Ratings and other User Generated Content created and published on third party websites.

TOP 2

• 50 views - http://pulse-beat.blogspot.com/2007/11/web-analytics-collectives-at-bcb5.html (Anil used the Web Reporting vs. Web Analysis slides, I guessed a number of views)
• 50 views - http://webanalyse-news.de/web-reporting-vs-webanalyse/ (Patrick posted about the Web Reporting vs. Web Analysis post and added commentary, I guessed a number of views)

I think it is fair to say that I probably not included all sources / measurement points, but I also think it is fair to say that I have enough to conclude on my hypothesis.


Conclusion:
What by first sight looked like the most popular page (What and how to measure Social Networking websites), using my traditional Web Analytics tool to measure Controlled On site Content - is actually not, by far, the most popular content that I produced. That would be; Web Reporting vs. Web Analysis.

2231 views - Web Reporting vs. Web Analysis
437 views - Controlled On site Content
1694 views - Controlled Off site Content
0 views - Uncontrolled On site Content
100 views - Uncontrolled Off site Content

879 views - What and how to measure Social Networking websites
879 views - Controlled On site Content
0 views - Controlled Off site Content
0 views - Uncontrolled On site Content
0 views - Uncontrolled Off site Content

It also becomes clear through my content quadrant categorization, that these choices depends very much on my vantage point – the same content can in essence qualify for all 4 quadrants (But much more about that later). Thanks to June Dershewitz for the Online Business Measurement Quadrant “vantage point” email dialogue.

The Online Business Measurement Quadrant

It is somewhat naive to believe that controlled on-site measurement is all that’s needed in this brave new world of User Generated Content (or if you want; Consumer Generated Media) – And I am happy to see that my good friend Avinash continuously preach that web analysis is not a thing you do in a silo.

I was trying to simplify this fact to a client not that long ago and ended up creating a structure that you might find useful. For the purpose of this post I named it the “Online Business Measurement Quadrant” .. .. which is kind of catchy.. heh? :-)


The Online Business Measurement Quadrant structure, try to shed some light on what we have to measure and in particular the very different environments we have to measure it in. The 4 quadrants represents:

• Controlled On site Content
• Controlled Off site Content
• Uncontrolled On site Content
• Uncontrolled Off site Content

Let me try to exemplify this by providing a set of content types.



Where I think it becomes obvious that even something as supposedly simple as collecting the activity becomes an obstacle. It is still fairly easy to collect good click stream data (Controlled On site Content), but I think you (as me) can see the difficulty in figuring out what to collect when somebody comments on your brand on some random blog (Uncontrolled Off site Content) - not even considering HOW we would want to analyze it and even more exciting HOW we would like to relate it to the data collected in other Online Business Measurement Quadrants.

Types of content

Controlled On site Content
Traditional marketing messages published on the corporate website, micro websites etc.

Controlled Off site Content
Traditional marketing messages published on third party website sites, widgets etc.

Uncontrolled On site Content
Brand commentary, Ratings and other User Generated Content published on the corporate website, micro websites etc.

Uncontrolled Off site Content
Brand commentary, Ratings and other User Generated Content created and published on third party websites.


JUST concluding that it is difficult moving beyond traditional web analytics measurement is hardly any help :-)
Let me provide you with a Random selection of vendors playing in each Online Business Measurement Quadrants and their typical measurement tool naming.



Find the list from the picture above expanded with links and markings (bold and a *) on some of the tools I use for this petite analytics blog.


Web Analytics - (Controlled On site Content)
(Measurement typically provided by your traditional enterprise Web Analytics Vendor)

- www.omniture.com
- www.webtrends.com
- www.indextools.com (*)
- ...

Audience Measurement - (Controlled Off site Content)
(Measurement provided by both large sample data vendors, but also channel specific vendors)

- www.nielsen-netratings.com
- www.comscore.com/
- www.feedburner.com (*)
- www.indextools.com (*)
- ...

Community Measurement - (Uncontrolled On site Content)
(Measurement typically provided by CMS, Blog or other community application)

- www.wordpress.org
- www.blogger.com (*)
- www.indextools.com (*)
- ...

Buzz Measurement - (Uncontrolled Off site Content)
(Measurement provided by a set of relatively new tools working specifically around Sentiment )

- www.umbrialistens.com

- www.buzzlogic.com
- www.buzzmetrics.com
- www.google.com/alerts (*)
- www.technorati.com (*)
- ...

N.B. I sent a request to http://www.scoutlabs.com/ for a private DEMO. Guys; if you are reading this, do send over that access as a Christmas Present :-)

As you can see, I added my traditional Web Analytics tool into the mix in other quadrants that the one we named “WEB” – and the reasoning is that enterprise analytics systems today CAN do a bit of data collection amongst other quadrants – just not as perfect as we would want to. A subjective and personal attitude on how much we can do would be visualized as follows:



This whole Online Business Measurement Quadrants chat is a work in progress – so please do email or comment so I can refine this to something even better :-)

Happy holidays to anybody actually reading this over Christmas..

Cheers
Dennis