We know our blog readers read our posts in a variety of ways. You come to us from desktops, mobile devices, social media, email, and apps.
We also know that these different portals to our blog can create different experiences. That's one reason that the HIV.gov team is always looking for ways to improve the content we produce.
One method for doing that is to see how many of our readers viewed our blog posts in their entirety. This is an important measure to determine if we are "engaging" our readers.
While blog topics and quality are the leading factors in determining if our audience reads an entire post, we wondered if the length of blog posts could also be a factor. So we set out to determine what percentage of visitors to our blog posts viewed the entire post to determine a uniform level of engagement.
1st Attempt: Heatmaps
We thought one way to answer our question might be to use a heatmapping tool. There are a number of options; we use Crazy Egg (requires a subscription).
Heatmaps can be very useful for tracking general website visitor trends and clicks on specific parts of webpages. The software produces visual representations of how far users scroll down on webpages.
We were looking for a measurement process that could be replicated for every blog post and by every team member. But heatmaps use color scales to represent popularity, and they don't take repeat visitors into account. That means they can be interpreted differently depending on who is doing the analyzing, because they don't provide a definitive number in relation to scroll-depth.
It also means that this tool didn't meet our specific needs—but, depending on what you are trying to learn, it might work well for you!
Tip: Heatmaps are a great tool to track clicks on specific webpage elements, like links and videos.
2nd Attempt: Scroll Tracking
On our second go-round, our analytics team discovered that we could use Google Tag Manager and Google Analytics to accurately and consistently measure how far users scrolled down. (Bonus: Both tools are free!)
We set the measurement thresholds at 25%, 50%, 75%, and 100% and waited for the metrics to roll in. We hoped they would tell us what percentage of users scrolled—and hopefully read—to the very end of our blog posts.
We discovered that we had more than 60% of visitors reaching the 75% mark, but we couldn't break through to the 100% threshold.
Again, we found some limitations to this approach. This technique didn't isolate scroll measurement to the blog text. It also included the header and sizeable footer sections of the website. But HIV.gov is designed in such a way that it isn't necessary to scroll all the way to the bottom of a webpage to read an entire blog post.
The other thing we discovered was that the metrics for very short blog posts that didn't require scrolling were misleading. Some brief blog posts only registered a 50% threshold—even though you could read the whole thing on one screen.
Measuring scroll depths can be helpful to see if users make it to different sections of webpages such as the footer; however, this was not the information we needed to determine if our audience was reading our blog posts to the end.
Tip: Google Tag Manager is the preferred deployment method for many web analytics tools and it allows for rapid testing and publishing of customized analytics.
3rd Attempt: Element Tracking – Success
There is a call to action at the bottom of each our our blog posta; it asks users to subscribe to HIV.gov's email updates. Our analytics team theorized that, if we could count the number of times this element appeared on unique visitors' screens, we could more accurately determine how many users view entire blog posts. To do this, the analytics team created a custom event in Google Tag Manager and Google Analytics to be triggered every time this call-to-action element appeared on a screen.
Tip: We recommend having a call to action on every page—not just for tracking purposes. It's important to help guide web visitors through their website visit, so they can find all the information and resources they need.
We're still collecting data, so we can't draw any solid conclusions yet. However, when we have enough information, we intend to share our findings and recommendations with the HIV.gov content team, federal partners, and guest blog contributors, so they can use it to write content at the optimal length.
For more tips on how to use metric insights to optimize your content, sign up for a Virtual Office Hours appointment and speak one-on-one with an HIV.gov digital specialist.