One of the advantages of using WordPress for my blog is it has a decent amount of built-in reporting. For example, WordPress ranks entries by the total number of page views since they were first posted. Unfortunately, the built-in reports are limited to last 7 Days, 30 Days, Quarter, Year, and All Time. This means it’s difficult to compare the popularity of an older post to a new one.
As an experiment, I extracted the page view data for all of my posts and ran a simple analysis of # page views / # weeks posted. This removes the obvious effect that most posts are only read in the first 2 weeks after they are posted. Not surprisingly, the top 10 posts didn’t really change much.
However, two 2008 posts (BI Predictions for 2008 and Love or Fear the Boss) moved into the top 10. The explanation for the BI predictions entry is straightforward: several other bloggers linked to it in their other predictions which drove traffic for a few more weeks. However, I’m unsure how to fully explain the traffic for the Boss entry. Some of the extra traffic likely comes from the fact that I referred back to it in a later post called Bad KPIs, bad behavior. However, it continues to get organic traffic every week; apparently people are searching for similar content.
While manipulating the data, I noticed something odd about the timing of blogs. Almost all of my entries were posted on Sunday but there are a small number posted on another day of the week. The mid-week blogs universally had a page views per week much less than average. While it’s possible these entries were less interesting than the others, that seems like too much of a coincidence. My guess is that people have gotten used to reading my blog at the beginning of the week and, if the entry shows up at some other time, it may no longer be in their feedreader when they get around to it.
What do you think – does the timing of blogs matter?
(P.S. It’s not lost on me that this entry doesn’t occur on Sun.)