Web analytics in action in higher ed
Students at Arizona State University today have access to a service-oriented financial assistance website that provides guidance on college planning. It offers cost calculators, links to scholarships and budget planning, and easy access to financial aid office staff. It’s an approach that works—and the result of using analytics to unearth a treasure trove of website insight.
“We used to be more policy focused, but after looking at analytics, we understood [finance] is a very important topic to students and visitors. We overhauled the entire site,’’ says Jill Andrews, executive director of enrollment services communications, which oversees ASU undergraduate and graduate communications for recruitment.
“We made it much more service-focused and user-friendly and did not just use [the financial aid page] as a place to communicate policy.”
But that’s not all officials discovered through web analytics. Using Google Analytics Premium (a paid version of the common tool), Andrews’ team discovered that Arizona State’s international page was generating a significant amount of traffic. “We know we’re a university that has a high number of international students enrolled …. [but] it surprised me a little to see the traffic coming in from across the world,’’ she says. “It changed how we built our international website and we now deliver some language-specific pages.”
Colleges and universities typically use the free version of Google Analytics to track how people interact with their websites, including where visitors are coming in from and the pages they’re visiting. Yet, industry observers say, not enough schools are taking the time to glean critical information from that data, such as why bounce rates (visitors leaving the site) might be high, why conversions are low (visitors not performing a desired action on a web page) and why more alumni aren’t making donations online.
Web analytics in higher education is “widespread, but in my opinion it’s not used in a very sophisticated manner,’’ says Alan Etkin, a research and analytics associate at consultancy Ruffalo Noel Levitz. Most colleges, he adds, will just “occasionally dip in to get some very basic metrics.”
Requests may come from a president or dean asking for traffic patterns, says Etkin, who is also senior analyst of web analytics at British Columbia Institute of Technology. “I’ve dealt with a few universities that are proactive in pulling numbers out of their tool on a monthly basis, but most of what I’ve seen is not that useful … so the way it’s being used isn’t providing the potential that it could.” Etkin pushes out analytics reports to about 20 different departments at his institution. But only more recently he has seen administrators making strategic decisions based on the information.
“They’re simply tracking activity on the website but not connecting it with conversions,” he says. For example, when there is a big event, the page that provides the basic information is separate from where students have to register, and that’s where Google Analytics often stops tracking, he says.
If administrators really want to understand the value of web visits, it is critical to track the whole trajectory of a visit—for example, starting from paid advertising and other dollars spent on marketing to the site visit made by prospective students, patterns of behavior and the actual application to enroll, Etkin says. “Most universities cannot track that.’’
But administrators at any college or university can use analytics to help make strategic decisions about content, advertising and web design with the right information and know-how.
Using analytics that’s offered
One problem with web analytics reporting is that officials outside of the web team or IT group may never see the data, often because they’re not interested or in many cases, they wouldn’t know how to interpret reports.
College administrators are no different than corporate leaders when it comes to the importance of understanding core website metrics. But in higher ed, administrators often don’t even know what kinds of analytics information will be valuable to them, Etkin says. Just comparing a website’s monthly visits to the same period in the prior year isn’t that meaningful, he adds. The tighter web analytics data is segmented, the more useful and actionable it is.
That’s the plan at Rollins College in Florida. Marketing and web officials have been tracking the top 10 pages people visit on the Rollins.edu site, says Jennifer DeWitt, associate director of interactive marketing and communications. In tandem with Linda Watson, web technology manager, the web team also determine which content to post based on the academic calendar, DeWitt says. Her team has been more closely examining the analytics over the past few years.
The plan is to personalize the data to track demographics. Recently, there was an alumni reunion and as people registered, Watson looked at who was coming by age group and gender. Now, she wants to know who specifically used the registration page so that going forward, it can be tailored to meet the needs of different age groups.
Watson also wants to parse data geographically, something she does frequently for the admissions department. Already she knows admissions gets a lot of pageviews from India, which she says is a distant second to the U.S. Although she hasn’t yet broken down pageviews by country, Watson is hoping to segment that next year.
Asking for more
A high bounce rate may indicate visitors can’t find what they’re looking for and that content needs to be adjusted.
More tightly honed data has helped the Medical College of Wisconsin in making more informed decisions about the content on its site, says Angela Nelson, senior director of development and alumni relations. She receives quarterly reports on the number of people who contribute online. One time, data (including visitors not following through on making a donation) indicated that people weren’t finding the right path to the donations page.
So the web team added a large “Donate Here” button to the center of the page, replacing a hard-to-find “click here” link in the text. The next step is to find out why visitors are clicking the button but not making a donation.
There have also been changes to the page that contains the form for registering for the annual alumni weekend. Users weren’t printing and mailing back the form, but analytics showed alumni were going to the page. Online registration was added—and more than 50 percent of registrants have chosen that option.
Keith Murphy, Medical College of Wisconsin’s director of web services, says he generates analytics reports for the president, vice president of public affairs and the board of trustees. The data show how many patients have used the website to pay a bill or make an appointment with a doctor; how many students have inquired about graduate programs or medical school; and how many people have filled out job applications.
Administrators mainly ask which pages are generating traffic, but what’s more important to know is whether visitors have opportunities to interact with the site, says Murphy. “If there’s not a call to action or engagement on their website, they’re missing an opportunity to create a relationship with a visitor.”
Once a department does that and increases the number of visitors, then they have to work on getting them “to buy something,” or in the case of a college, “figure out what we want them to do.” That goal might be getting more students to enroll, having more people to apply for jobs, or making more online appointments with a doctor, since the college offers patient care and is interested in enhancing those areas.
Understanding the art of analytics
Administrators should not treat analytics information “like it’s accounting,’’ says Murphy. “It’s more of an art than science.”
It also involves using the data to interpret trends over time. “How many hits you got yesterday doesn’t mean anything,” he says. “You want to say, ‘Was the number of visitors yesterday more than a week ago or a year ago, and is it going in the right direction? Did they do something on the site that I want them to?’” This will help administrators delve deeper into what Murphy wants them to think about: Are they serving their constituents through their website, and how do they know that if constituents aren’t engaging with them?
Etkin uses end-to-end tracking to determine revenue generated when, for example, British Columbia Institute of Technology spends $10,000 on advertising. He also can track the number of people who apply online and are accepted to the school after attending a certain admissions information session—and run conversion rate-based calculations to show the value of every attendee.
“We can use that number to calculate what the investment should be—how much we should spend on advertising the information session,” he says. Etkin also acknowledges analytics can be daunting. And even at his institution, where he has begun including revenue data in reports, the responses have been mixed. “One of the clear lessons to me is the more you measure, the less you get done. I think that’s a common mistake,’’ Etkin says. “Having said that, there are real nuggets to be found.”
The trick, he believes, is to make sure you connect your conversion, get your Google Analytics code into where the registrations are taking place; where your school is taking in money.
Murphy agrees. “Analytics helps us figure out what is the weakest link in the chain.”
As Watson at Rollins notes, “The bottom line for any college is to get people in seats and giving. If we don’t do those two things, we’re not here.”
Esther Shein is a Framingham, Massachusetts-based writer.