The Rise of Analytics-Driven Decisions

The Rise of Analytics-Driven Decisions

Relying on web analytics in decision making

Have you heard about the analytics revolution in higher education? Ready or not, it’s coming to your institution—if it isn’t already there. Whether you work in an academic, business, IT, marketing, or web office, the data-driven movement is slowly but surely making its way in to the hearts and minds of top executives faced with serious strategic and financial challenges.

Think this is just wishful thinking from the higher education online analytics evangelist I’ve become over the past two years?
Educause begs to differ.

In October 2011, Educause launched a new initiative, using a Bill & Melinda Gates Foundation grant, to advance analytics at U.S. colleges and universities. It shares some common objectives with the “higher education online analytics revolution” I put in motion two years ago with a series of columns, a few conference presentations, several benchmarking surveys, and, more recently, the launch of a four-week online course on web analytics.

By developing education, resources, and professional development programs and conducting a major benchmarking study of the state of analytics in higher education, Educause is going to make sure the analytics issue finally gets the attention it deserves in higher education.

When I launched the first online survey about the state of web and social media analytics in higher education in May 2010, my goal was to find out if a change in attitudes toward analytics was necessary.

This year, I conducted its third edition a bit earlier, to share the results during my UBTech conference presentation this month in Las Vegas.

Completed by 344 professionals working mainly in the marketing, communication, and web offices of their institutions, this voluntary online survey about the state of web and social media analytics was filled out from March 30 to April 18, 2012. It was publicized online only through several different channels (email listservs, email newsletters, blogs, Twitter, Facebook, and other professional networking websites, etc.). As a consequence, it doesn’t rely on a scientifically determined dataset. However, the resulting sample of surveyed institutions seems fairly representative of the Carnegie Classification breakdown. It is also consistent enough with the samples from the 2010 and 2011 surveys to allow comparisons.

Survey Highlights

Web and social media data tracking in higher education has kept increasing over the past two years, from 95 percent in 2010 to 97 percent this year. During the same period, Google Analytics (GA), the free yet very powerful analytics application, has consolidated its dominance of the web analytics solution landscape in higher education by increasing its market share from 92 percent of the institutions tracking web data in 2010 to 97 percent in 2012.

The 2012 survey results also reflect the incredible increase in social media analytics usage by institutions. While 36 percent used Facebook Insights and 24 percent used YouTube Insight in 2010, 62 percent and 40 percent, respectively, now rely on these platforms to track the performance of social media activities.

Colleges and universities have also started to allocate more human resources to analytics, as well. Two years ago, only 25 percent of respondents indicated they spent more than two hours per week on analytics. That proportion has steadily grown to reach 38 percent this year. More than a quarter (27 percent) of 2012 survey respondents still spend only two to five hours per week, but a very small number of institutions, including Ithaca College (N.Y.), have started to hire full-time analysts. The needle is moving very slowly, but in the right direction.

Aside from other data points (which can be reviewed in the full “2012 State of Online Analytics in Higher Ed” report available online), this year’s survey invited respondents to share analytics success stories. One in five respondents did. Here is a very small sample of the most inspiring ones.

Website Overhaul

Almost three quarters (72 percent) of the 2012 survey respondents depend on analytics to improve their website. At Lemoyne College (N.Y.), analytics data was used to vividly illustrate the main shortcomings of the institutional website, namely the reliance on a static—and not interactive—homepage banner image and its lack of mobile-friendliness.

A thorough analysis of the visits and page views from mobile users, as well as the homepage bounce rates, was performed. “Because I was able to illustrate both trends with hard data, I was able to successfully make the case for not only a redesign, but a mobile friendly site and the use of responsive techniques,” recalls Michelle Tarby, director of web services.

At Eastern Kentucky University, the same mobile traffic metrics provided justification for the fully responsive redesigned website of the IT department that was launched in December 2011. “In GA, we looked at the increase of usage of the site, and the number of mobile users. As both increased, we knew we needed to meet the needs of our users,” explains Tammy Cornett, EKU coordinator of IT communications and development.

A Voice for End Users

In other cases, analytics are critical to inform design decisions, ensuring websites better serve target audiences or even institutional goals. Over the past few years, web analytics have become the best friends of web professionals trying to ensure design decisions benefit the silent majority of end users.

At Johns Hopkins University (Md.), web analytics helped the Sheridan Libraries initiate a radical web purge in 2010. “Doing click tracking and bounce rate analysis allowed me to cut the website from hundreds of pages to three,” says Sean Hannan, senior web developer for JHU Libraries. According to Hannan, the website has now transformed into a site that quickly and efficiently sends visitors to the web resources they need.

Ben Greeley, the web manager at Colby College (Maine), adopted a similar approach to steer departmental website stakeholders in the right direction. “I met with the departments and analyzed what pages/sections were getting traffic and which ones weren’t,” he recalls. Then, each department could compare the trends with the department goals and resources available to maintain the website.

By giving a “voice” to the end users, web analytics can solve internal politics nightmares … before anybody gets hurt.
Whether it’s a campus department that is requesting web content pertaining to two very different areas be hosted on a single website for the convenience of its own staff or a top campus executive who wants a direct link to his or her unit website on the homepage global navigation menu, web analytics has saved the day at several institutions that shall remain nameless to protect the innocent—and the guilty.

Analytics for Enrollment

At Washtenaw Community College (Mich.), web analytics even helped grow enrollment in online, blended, and weekend classes. “There was a host of confusing information on the website that resulted in student confusion, lower than desired registration, and high levels of support calls,” remembers Christopher Billick, director of web services. Through a systematic analysis of the website internal search terms, it became obvious that many students were not using the branded terms—“College on Demand” and FlexEd—that had been created to market these classes, but rather more intuitive and simple keywords like “online classes.”

In the case of the University at Buffalo (N.Y.), a monthly analysis of online donations and their referral sources revealed that some UB schools did not provide a direct link on their website to the institution’s online giving form. “We met face-to-face with most schools and showed them the stats we had. The majority were shocked enough to immediately add the link,” shares Kara Sweet, web communications specialist in the office of development communications.

At the University of Chicago, You-Tube analytics data supported the case to rename an online campus tour video that was originally called “A Student-Led Introduction.” Dave Pickett, social media editor, delved into YouTube analytics. As it turns out, none of the search traffic to the video was coming from the phrase “University of Chicago campus tour,” he says.

Something wasn’t right, as this phrase was indeed the top UC search term on YouTube. By simply uploading a second instance of the same video­—one that was better labeled­—the tour video got twice as many viewers in half the time than the original without any additional promotional efforts.

As all of these examples demonstrate, time spent on analytics is indeed time well spent.


Advertisement