How does Google Analytics Help?
In our last post we asserted that using Google Analytics (GA) isn’t about enumerating visits, device types or operating systems, it’s about gathering the data to understand if a site’s objectives are being met and to what degree.
The first step is to identify the outcomes that the overall site or specific pages are intended to achieve and determine what measures would let you recognise success.
Higher education institution websites exhibit two specific complexities that makes identifying outcomes more difficult than for most ecommerce or other commercial websites.
Many (most?) universities and colleges have large numbers of autonomous ‘sub-sites’ and each site may have quite distinct goals. A main home page might focus on student recruitment, while research centre sites could be promoting their work, faculty and events. The second peculiarity is that site content addresses disparate sets of audiences.
The combination of these factors means that for higher education websites goals, objectives and outcomes need to be very specific. This post discusses five cases to understand if the intended outcomes are being met.
The second step is to ensure that GA collects data that can actually measure the goals, outcomes or user tasks you've selected in step one.
We show how GA combined with Google Tag Manager (GTM) can collect outcome data to measure success through ‘use cases’ that match typical higher education website objectives. The collected data, in turn, lets evidence-based changes be made to page content or design as appropriate.
Before plunging into using GA/GTM, in anger, we have four suggestions:
- You should subscribe to Simo Ahava’s blog. The blog provides practical examples of GA and GTM use that can be implemented on higher education websites.
- If you distribute content via social media, email or post on LinkedIn, Medium or elsewhere use a campaign tracking tool such as, Campaign URL Builder to encode content links for subsequent identification in GA.
- You will find analysing the data passed to GA easier if you understand how to set up Custom Segments or you know how to write custom reports in GA or better yet, use Google Data Studio.
Each ‘case study’ is intended to illustrate measuring website goals to understand how well objectives are being met or gather longer-term data to inform website re-design exercises.
Case 1: We publish plenty of content. Is anyone reading it?
Between marketing and communications campaigns, blogs, online newspapers, media releases, student-generated content and research publications large amounts of new ‘content’ are added to higher education websites daily. Each of these contributions represents an investment in time and money in preparing material with specific intents, audiences and outcomes.
How do you know if those goals are being achieved and an appropriate 'return on investment' is being achieved? Typically, understanding content popularity involves recording ‘hits’, pageviews, bounce rates or average times on page. These are macro engagement measures and fail to provide actionable feedback about specific content or allow inferences about general content engagement levels.
A solution: GTM provides a trigger activated by how far down a content page a visitor reads or scrolls: the scroll depth trigger. Twinned with an appropriately configured tag, you can measure whether visitors are scrolling halfway or all the way to the end of blog posts or articles and at every point in between. And, by examining the data for all content being ‘read’ you can determine rules of thumb for engaging content length.
Case 2: We list ‘interesting’ links on our pages, but does anyone use them?
Website content rarely exists in isolation and depending on its intent, authors often link to other material, authors or sites that they believe are relevant to their audiences.
However, web page space should be allocated to content of maximum value to its audience(s). If links aren’t being clicked, why not replace them with content of higher value?
A solution: GTM provides a ready mechanism to record clicks on any outbound link on any website page. The solution to gathering the relevant data combines a variable to capture the hostname (“URL”) of the clicked link with a trigger that fires when the hostname does not match site’s domain name. Step-by-step instructions on setting up outbound link recording are in the post: #GTMtips: Track Outbound Links In GTM V2.
Case 3: We’ve placed forms on our site, but our sign-up rates seem low, what’s happening?
Newsletters, event notifications and other content is often made available through email sign up forms. And, the consensus seems to be that less is more: the fewer the boxes to be completed the higher the probability that visitors will supply their details.
Email sign-up forms typically link to email campaign management systems providing subscription reports. Understanding how many visitors fail to complete sign-up, isn’t available from campaign management systems, but is key to identifying issues with form design, requesting too much data or not offering sufficient rewards for subscribing. Google Tag Manager to the rescue – albeit with a few caveats.
For the purposes of this case study we assume that a page contains a single sign-up form (for example: first name, last name and email address) using basic HTML formatting. Google doesn’t want personally identifiable information (PII) passed to GA, so user data entered in the form fields isl NOT be passed to GA – but a record of any activity in these fields is.
Case 4: We’ve made sure important information is available as PDFs. Does anyone read them?
In our blog post Higher Education's Website Content Epidemic we highlighted the huge number of PDF documents on higher education websites: 1.7 million files across the websites of Canada’s U15 universities.
There are many reasons (excuses?) PDFs are used on university and college websites: they create ‘permanent’ records, are easily uploaded and readily open in browsers for reading. The downsides are that few PDFs are accessible (almost none of the PDFs detected in our higher education website audits are PDF/UA) and we speculate that only a fraction of these files is ever read.
However, one need speculate no longer, as implementing PDF link tracking allows us to understand PDF file interactions and accumulate data on document popularity. Popular documents are surely candidates for conversion to web pages? And, the least popular, candidates for elimination?
A solution: GTM allows triggers to be set based on CSS selectors, the page elements to which styling is applied. GTM can be configured to recognise links: <a href>https://www.exampleu.ac.nz/calendar.pdf </a> And, together with text and pattern matching, using regular expressions specific types of links or link content can be selected. In this example, pattern matching looks for ".pdf".
Simo Ahava’s blog works through the specifics of how to implement this capability in GTM and, thus, being able to assess the merits of using PDFs rather than HTML-based content.
Case 5: We’ve made sure that relevant email addresses are on our site, is anyone clicking on them to get in touch?
In our blog post, Why Are There Email Addresses All Over Your University Website? we examined the prevalence of email addresses on higher education websites. About 40% of the 40,000 web pages we audited included ‘mailto: links’ allowing email contact.
A key question about including email addresses across university and college websites is: is anyone using them? The answer is important, as our website audits show that email addresses contain errors, can be outdated as staff and faculty change and the types of addresses used often diverge from institutional policy preferences, employing webmail rather than official university addresses. Moreover, embedding email address links across websites increases the maintenance burden on already scare resources.
If mailto: link use is low, why not direct visitors to centrally-maintained staff and faculty directories, instead?
A solution: Gathering data on mailto: link usage follows the same approach as finding links, by recognising relevant CSS selectors, in this case mailto:. GTM can recognise links, such as: href="mailto:firstname.lastname@example.org" and pass a record of each click to GA. Collecting email address usage data can identify high demand individuals, departments or distribution lists, as well as establish cost-benefit cases for centralising access to contact details.
As we summarised in our last article, Why Higher Education Websites Need Google Analytics, higher education institutions have enthusiastically implemented Google Analytics (at least on their main websites), but have been slow to install Google Tag Manager. That’s a shame. As the five case studies in this article illustrate GTM lets marketing, communications and other groups capture invaluable data about how their websites and content actually perform.
In a world of scarce resources and competition for attention, it makes sense to do more of the things that yield the right outcomes and less that yield the wrong or no outcome. But, you can’t separate wrong from right if you aren’t collecting any data in the first place.