My Photo


  • Clicky Web Analytics

Your email address:

Powered by FeedBlitz

« Don’t Settle for Leftover Turkey When it comes to Web Measurement | Main | More on Analytic Reporting Models »


I do a lot of reporting like this and have found myself using graphs with bullet points of analysis to explain the changes. It often ends up being very wordy because of all the factors involved.

I like your usage of the "% Change Explained" field. As a web analyst, it is easy to get caught up in looking for all the reasons that a metric has changed, but it is often not worth the time to find the last 10%.

Thanks Gary, This is very helpful. I have to agree with Ben, graphs and bullet points are standard. Bloated reports, too ;) One question and one recommendation.
a) Any way to get a sample spreadsheet?
b) I was reading 90% of the post thinking "same old, same old..." Why hide all the good stuff in the end? Putting the first graph in front would have hooked we from the first second. Then you could have gone into details.
Regardless. Great post. I will have a try at creating a similar chart :) Oh, the beauty of correlation.

Great post! I really liked the Traffic Model that you've created . It makes the report much easier to read and understand then if you show endless charts that explain the changes.

It would be great if I could see an example of what you've created. Any chance you could send it to me?

Great article distinguishing the real information that can and should be gained. Educating the consumer is such a large part of the reporting.
I also would be interested in receiving or hearing about getting that data out of the server stats (or client side?) and your preferred analytical package that provides that (by article or email)... and of course if it is not too much to ask that 'fairly complicated' macro spreadsheet. I guess I believe in Father Christmas. Have a good holiday break if you take one.

When you speak of models do you mean probabilistic modeling? How do you factor in one off events (media event, competitor actions, ect.) that might be either directly or indirectly (through your explanatory variables) influencing your visits?

The comments to this entry are closed.