One of the first goals of any successful blog is to find ways of driving traffic to one’s site. It is all about getting attention. One of the most obvious initial lessons about blogging is that the higher the cumulative number of posts, the more page views one generates.
The chart above shows an exponential relationship between daily page views and my site’s cumulative number of posts. Regression analysis suggests that the total posts I cumulatively generate explains about 37% of the data for daily page views.
If one removes outliers of more than 200 page views in a single day, the equation below describes nearly 40% of the data.
The bottom line is that not only are one’s page views directly associated with the number of posts one generates, but that there is a predictive relationship between the two numbers that can be described by an exponential curve.
What this means in practice, is that most blogs will start slowly and then begin generating momentum.
This result is intuitive because the more posts that one has on the internet, the more opportunities one has for readers to stumble across one’s site.
Of course, such an exponential increase can not be sustained forever.
My view is that like any product, a blog’s page views follow an S-curve. Readership slowly and steadily increases, then it accelerates until it reaches a critical mass. It then increases slowly again as it approaches a theoretical saturation point.
In the next installment, I will discuss the next lesson: leveraging the media to drive traffic to your site.
Real analytics! While it seems intuitive it seems you have proven the value of posting. Thanks for sharing.
I’ve noticed you’ve been posting a lot more since my first blogging article. 😉
One of the most obvious initial lessons about blogging is that the cumulative number of posts, the more page views one generates.
I’ve thought about this as well. The other metric I’ve thought about is hits per day vs posts per day. I wonder if more = more there as well.
I’ve always thought that at some point, your reader will come to the site and read what’s there, be it 1 post or 3. However, over time, readers will be conditioned to the number of posts ad begin to check more often. As such, I have am beginning to think about timing my posts to a certain time of day. Condition the reader to come back for more–more often.
Not sure if the correlation between hits and posts really gets at the true drivers?
I think the ultimate driver of personal blog success is primarily to either get ahead of the curve on either info or insight (i.e. offer it first), or have the “Howard Stern Factor” going where people are simply fascinated to read the wild comments a blog’s author or its commenters make.
Of our four emotional needs: Variety, Significance, Connection, and Growth, I think a blog needs to be a high-level deliverer of at least two of those needs to be “good”, three to be “Great!”, and four to be “Oustanding!” Hit all four needs at a high level and you’ll really get people hooked.
There is definitely a lot more to driving traffic to one’s site than cumulative posts and posts per day. These two are just the easiest ones to measure.
Don’t worry, I have 10+ blogs coming on this topic in the next few weeks.
Hey, wouldn’t know you at all if I didn’t think there was more to it! 🙂 In my job I’m always pairing the quantitative with the qualitative, so I guess I was just anxious for the “other half”!
Looking forward to reading the additional posts.
I also looked at posts per day. If I include the outliers over 200 hits and do a multiple regression that include cumulative posts and posts per day, the curve is still an exponential one. This curve explains 49.2% of the data. If I remove the outliers, the resulting curve explains 51.6% of the data.
The reason I trust this result a little bit less is for two reasons. First, when I had a successful day, I would typically start writing more posts to take advantage of the increased traffic. In essence, page views were driving posts per day for me to some extent.
Second, the regression suggests that the effect of more posts increases exponentially as well. I think this is true for one you are talking about 1 or 2 posts. However, I believe that this effect should start to decline after 3 posts. The data does not show that.
For instance, my equation predicts that I should get 161 hits today with 1 post. I think this is a likely and accurate outcome (Remind me tomorrow to tell you what the actual number is). However, the same equation predicts that I will get 492 pages views with 3 posts. To be blunt, I never had this many page views in the history of this blog.
I also looked at timing (i.e., days of the week) and it turned out that it was not statistically significant. However, I think there is something to time of day. Anecdotally, my hits seem to come in three waves. Early morning, around lunch, and around dinner time.
I would love to analyze this data in more detail. The problem with WordPress is that I would have to manually enter these times for every article I’ve written, which is a huge time sink.
In essence, page views were driving posts per day for me to some extent.
Interesting. Not me.
I also looked at timing (i.e., days of the week) and it turned out that it was not statistically significant.
Well, THAT is goo news. I have avoided posting on weekends and late at night for fear of missing the audience.
Another massive help is development of the community. Comments drive comments!
Absolutely. I will cover this in more depth as this series continues.
One entrepreneur in Silicon Valley has also told me that odd things like changing the color of a website can drive pages views up by 20% on a given day. Humans, apparently, are hard-wired to respond to various stimuli in different and often unpredictable ways. I am going to experiment with things like this as well from time to time.
Now I’m up to post #3. The algorithm predicts I should now hit 507 page views today (460 if you remove outliers over 200).
By the way, with that second post, my algorithm now projects 286 page views for today.
As you will likely see, the algorithm starts to fail at higher numbers of posts per day.