Tracking people who use the Library's book list blog posts

No comments

Someone asked me recently, "Can tell if people who look at the book list blog posts actually click on the books and borrow them?" A good question, and as I looked into it I thought it might be interesting to share some insights. Spoilers: the answer is yes... and no!

The blog posts I'm talking about are the book lists that look like the one in the image above. These posts appear as a link on the home page of the website and they are also emailed to 426 subscribers.  When someone looks at one of these posts ideally they will click on the Get it! button for books they like, which links to the book in the library catalogue.  They can subsequently reserve the book if they want.

When it comes to tracking user behaviour in this user flow we can get half way.  I have placed some tracking code in the button links so that I can tell from Google Analytics when someone visits the library catalogue from one of these links.  From this data I can tell which posts get the most engagement. I can't tell, however, whether people subsequently reserve or borrow the books. This would require adding further tracking code to Encore, which is difficult.

While I haven't gone into great depth looking into this, here is a snapshot of the data with, what I think, are some interesting observations.

Overview of blog post traffic to the library catalogue

This chart comes from Google Analytics and shows an overview of the traffic to Encore from blog posts, covering the period from 1 January to 31 October this year. First some raw numbers:
There were 207 unique users who click on links from the blog posts. Those 207 users initiated 773 sessions - while it doesn't correspond exactly to the number of links clicked, for our purposes, you can think of these 773 sessions as being 773 books from the blog posts that users clicked on.

You can drill down into how much engagement each of the different posts received. I've saved the data in a Google Sheet called Blog Post Acquisitions, which can be accessed from the Library's Google account. This spreadsheet shows the same numbers as the chart above for each separate post, so you can check out how the different lists performed.

If you look at the chart above there are 2 more columns that caught my eye in these numbers.

Firstly, Pages per Session. For the blog posts this averages out at 6.46 pages per session. This means that every time someone clicks on a link from a blog post to the catalogue they then visit a further 5 or 6 pages – that is, they do some further searching in the catalogue. This is an average, so some people visit more and some less of course. By itself this number is not that meaningful. The interesting thing for me is that the Pages per Session for traffic from blog posts is higher than the average for the catalogue generally (4.79). 

This leads to the second number, Average Session Duration. It makes logical sense that if you view more pages in a session you will spend longer on the website. In this case blog post visitors spend over 7 minutes on the catalogue (7:24), which is more than double the site average of just over 31/2 minutes (3:37).

The traffic that is generated by these reading lists only makes up a small portion of the overall site visitors, however, these people are a more engaged cohort.

The final thing I noticed when looking at the shape of the graph in the image is how it tells the story of our year. We begin the year at a high level. It then starts to dip as COVID starts to impact the community. There's a bump in April, right after we close the library, where we have people coming to the website to find out what's happening. Then there's a low period right through to August, before the traffic starts to rise again to similar levels as the start of the year.  Throughout that low period from May to August our blog posts were only including links to ebooks and eaudiobooks – RBdigital, Overdrive and Borrowbox – while we tried to support our members through our closure.

Martin

No comments :