BiblioCommons Blog

Using A/B Testing to Drive More Blog Feed Clicks

Written by Eliana Franco | Dec 7, 2023 3:00:00 PM

One of the challenges of maintaining a public library website is ensuring that it continuously appeals to the needs and preferences of your online patrons. You might align with user expectations at the start of your website’s lifecycle, but people change, and so do the technologies and standards underpinning the web. If you don’t continuously stay on top of those shifts on the web, you will not exceed your user reach and engagement goals, especially through the long term.

In this blog post, we’ll walk you through one example of how our team actively studies and incorporates new web design ideas and techniques into BiblioWeb, our integrated content management system (CMS) and webpage builder for public libraries. Our team wanted to see what they could do to drive more user traffic to our Partner Libraries’ blogs. By leveraging data and carrying out A/B tests, our team identified key design changes that can increase blog clickthrough rates (CTRs) by 180%.

 

 

Transform Your Library Website into a Vibrant Patron Engagement Platform

Demystify web design, UX, security, accessibility, and more with our complete guide and checklists!

 

 

The Problem: Are Patrons Finding Interesting Blog Content?

When investing in blog content, public libraries want to see their older posts continue driving net-new views, especially when those posts are no longer on page one. One way to drive new users to that content is to promote it on other blog posts and, in turn, encourage discovery.

The BiblioCommons team wanted to see if there were any opportunities to drive more of that discovery within the blogging capabilities of BiblioWeb. Our team found that when a patron read a blog post, it was effortful for them to find other related blog content. The reason for this friction was that the only promoted content was textual (i.e., it used taxonomy terms, tags, and posts from the same author).

 

Interested in learning more about BiblioWeb and BiblioEmail?

 

 

In contrast, when looking at other blog publishers, the team saw that they leaned a lot more on visual assets to promote related content. To unpack the strategies of the other blogs and to develop a better idea of what made them different from BiblioWeb, the team did two things: First, they reviewed the heat maps of BiblioWeb’s existing blog post functionality. This process enabled the team to see how patrons were interacting with our partner libraries’ blogs and to see which modules or links they’d click on to see related content. Second, the team conducted a deep comparator analysis to see how others (such as Medium, Vox, CBC, and The Atlantic) organize their related blog feeds.

With these insights in hand, the team explored how they could leverage using multiple content feeds as well as greater visual presentation and mobile visibility to drive more blog traffic.

The original set up only had one text-based feed, which showed more blog posts by the same author and was placed on the sidebar. However, this feed was only visible to desktop users, hence limiting the reach potential for recommended content.

The new setup (which is now live on BiblioWeb) is vastly different. Not only does the new setup leverage multiple feeds, but the team created new types of feeds. The team wanted to serve recommendations from different sources, such as taxonomy terms and audience taxonomies. This approach automatically recommends content from a variety of angles, be it more from the same author, specific tags, genres, or other identifiers. Moreover, two of the three new feeds were also visible to mobile users, thereby expanding reach to viewers who were otherwise excluded.

A/B Tests: Putting Ideas to Scrutiny

Though they got an idea of what feeds to add, the BiblioTeam wanted to experiment with how to best present that new information. So, they carried out a series of A/B tests with the new features in order to narrow down an optimal set up for use in the production version of BiblioWeb.

 

 

Drive more item circulation from your collections by
integrating your catalog and website experiences.

 

 

Test 1: Swapping the Positions of the Feeds

For the first step, the team went in with the hypothesis that changing the position of the ‘More by [author]’ feed with ‘Similar Posts’ feed would increase CTRs.

Compared to the version with the original configuration, the new arrangement (Version A) drove more conversion. As Version A was the winning design, the team built on it in subsequent tests.

 

Test 2: Swapping the Position of Card Feeds and the Comment Box

Building on the first test, the team decided to see if positioning the blog card feeds above the Comment block would increase CTRs. For this test, the team built on the winner of the previous test (Version A) by placing the blog card feed above the Comment block. This new page (Version B)  was then tested against Version A. Version B drove more conversions than Version A.

 

Test 3: Comparing the Winner of Test 2 to Production

For the final test, the team hypothesized that their collective changes on the page (Test 2 - Version B) would generate more conversions compared to the existing page layout on BiblioWeb’s production build.

The blog feed on the production version generated a CTR of 1.0%, while the blog feed configuration on Version B produced a CTR of 2.8%. Thus, the team was pleased to see that by placing the additional card feeds above the comment box, they produced a 180% improvement in the CTR on blog feeds compared to the production version.

 

Test 4: Mobile

For this test, the team hypothesized that a tiled display of blog cards on mobile (Version B) would drive a higher CTR compared to a display with full-sized cards (Version A). Interestingly, the test showed that the version with full-sized cards drove a higher CTR, thus proving the hypothesis wrong.


Implementation Steps

Before rolling out these changes to all BiblioWeb-powered libraries, the BiblioCommons team revisited the user experience (UX) of the card feeds above the Comments block. While moving it lower can boost CTR, they also know that user-generated content (UGC) like comments, reviews, etc., play a major role in driving patron engagement on our partner libraries’ websites. So, they looked for ways of building a connection between blog post contents and the Comments block.

Next, the BiblioTeam further developed the winning display (Version B) and, in turn, released the build to all of our Partner Libraries. The BiblioTeam did this test with the help of a few of our Partner Libraries  (Arapahoe Libraries, Chicago Public Library, Christchurch Libraries, Cincinnati and Hamilton County Public Library, Kent District Library, Pima County Public Library, Sno-Isle Libraries), but every BiblioWeb-powered library benefitted from this work.


 

 

BiblioWeb offers a ready-made public library website platform and rich tools designed drive patron engagement.

 

 

 

Why BiblioCommons?

A great aspect of this study is that it encapsulates only one among many enhancements our engineering and product teams carry out for our Partner Libraries . Even a comparatively small layout change requires considerable testing and reliance on data to deliver quality and positive impact. You can imagine how much capacity it takes to implement this level of work across every feature of your digital library stack.

By leveraging BiblioWeb, Partner Libraries benefit from the support and direction of our entire web development team. Our Partner Libraries don’t have to build a team from scratch or worry about raising necessary infrastructure to benefit from comprehensive web development work. Instead, they draw on the BiblioTeam’s work — including usability testing and data — to gain rich toolsets and capabilities to achieve their respective goals.

For more insights from our Partner Libraries:

 

Sign up to receive the latest in public library topics delivered straight to your inbox!