June 2, 2014

Traffic Drop During a Blogging Break

English: a chart to describe the search engine...
English: a chart to describe the search engine market (Photo credit: Wikipedia)
As regular readers know, I was away for a bit recently. Specifically, I did not post between May 15 and May 24. When a blog suddenly stops posting, traffic plummets, right? I figured it might be fun, at least for the bloggers among you, to take a look at some Google Analytics data and see what happened.

I selected two weeks for comparison:
  1. May 7 through May 14, the week before my break
  2. May 15 to May 22, my first week without writing any new posts
Was there a significant difference in traffic between these weeks? Yes. Between May 7 and May 14, Google Analytics reports 22,973 pageviews. Between May 15 and May 22, this dropped to 15,472 pageviews. Clearly, there was a drop. It was not as dramatic a drop as I might have predicted, but it was certainly a drop.

My initial hunch as to why the drop was not larger had to do with my social media strategy. I have been using Buffer to distribute content on Twitter, Google+, and Facebook. When I am posting new content, it goes out through Buffer, but I set it up to send select older posts out while I was away. That means that the frequency of social media activity was constant across these weeks. Could this have helped to mitigate against a larger drop in traffic?

If that was the case (i.e., if this social media approach helped minimize what would have been a larger drop in traffic), one might expect social media traffic to make up a larger proportion of overall traffic during the week I was away. This was not the case. In comparing the percentage of traffic coming from Twitter, Google+, and Facebook, the differences were too small to be meaningful.

When we ask ourselves why a blogger who takes a break experiences a drop in traffic, we suspect two culprits: direct traffic and search engine traffic. A drop in direct traffic is what we expect because RSS activity stops completely. Those who use RSS readers to follow a blog, subscribe to a blog via email (which is delivered through RSS), or visit a blog from links on other blogs (also typically delivered through RSS) see that there is no new content and have less of a reason to visit. Similarly, new posts generate search engine traffic so it makes sense that the lack of new posts would be associated with a drop in search engine traffic.

In looking at the Google Analytics data, direct traffic did decline. It dropped from making up 27.33% of the total traffic down to 17.95%. Unexpectedly, search engine traffic actually seems to have increased. Before the break, it made up 40.88% of the traffic; during the break, it made up 51.33%.

What is going on here? The math whizzes among you have undoubtedly spotted the problem. These percentages reflect the proportion of traffic from each source to total traffic. As such, they are quite misleading. For example, the apparent increase in the percentage of traffic coming from search engines does not necessarily mean more search engine traffic; this number is affected by all other sources of traffic. It may be that the proportion of traffic from search engines increased because direct traffic declined. To do what I want to do, I can't use these percentages at all. I need to use the raw numbers.

What do I find when I use raw numbers? There was a small decline in search engine traffic and a large decline in direct traffic. In fact, the direct traffic during the break was less than half what it was the week prior to the break. What this tells me is that the primary reason my traffic dropped during the break had to do with the drop in direct traffic. Other sources of traffic declined slightly, but this was the big difference.

An interesting question to ponder for future breaks is whether a significant increase in social media utilization during a break could help to reduce the drop in overall traffic. Suppose I had doubled my social media activity during the break. Would that have resulted in there being less of a drop in traffic? I do not have data that can answer that question, but my guess is that it might help a little but would be unlikely to make up the difference. The thing I am starting to learn about social media is that the impact on blog traffic is quite trivial compared to the time and effort that goes into it. But that will have to wait for another post.