IT Trends

When Students Whine About WiFi on Twitter

The University of Georgia and Arizona State have turned to Twitter feeds to monitor just how satisfied students are with campus wireless. Here's what they've learned.

• By Dian Schaffhauser
• 05/11/17

Mean tweets aren't just a feature on Jimmy Kimmel Live! — they permeate the feeds of college students across America. And the primary target of their wrath seems to be campus WiFi:

"Here's the deal... You can either give me functional WiFi or you can give me back a portion of my tuition. Which one is it gonna be?"

"What if this is all some kind of social experiment to see what happens when you take WiFi away from college students?"

"How to academically ruin a college student: shut off their WiFi on a Sunday."

Two institutions are using those kinds of disparaging tweets from their community of students as inspiration for improving delivery of wireless networking. At the University of Georgia, mean tweets sparked an initiative that has resulted in a major upgrade of residential WiFi. And Arizona State University is using its students' social WiFi complaints to reach out daily and publicize its 24/7/365 help center.

Here's what both schools have learned about how to cope with WiFi whining.

Twitter Isn't the Best Medium for Two-Way Communication

If your primary intent is to pinpoint the precise WiFi problems via the same mechanism students used to grumble about it, you may get frustrated. "When we would try to respond to students or answer their questions, sometimes we would have students give really detailed information as we continued the conversation. Other times they would complain about something and then not ever respond back," said Kerri Testement, U Georgia's senior public relations coordinator for Enterprise Information Technology Services (EITS). Part that is a result of when the tweets come in. "They don't want to call somebody — especially if they're mean-tweeting at 2 o'clock in the morning and our help desk is closed at that time," she observed.

Similarly, at Arizona State, when a tweet surfaces mentioning "ASU" and "WiFi," the university's help center responds as quickly as possible with a suggestion that the student call a number to speak with a support person. Hardly anybody does, noted Eric Dover, director of the help center.

But then that's not necessarily the point of the response, he added. "Because it's so public, that message is out there — their friends see that you responded. Others that may be interested in the university see that engagement and the response [as] a sincere interest in trying to help them get the best experience possible. So, it's two-faceted: It's helping that student, but it's also going, 'Hey, we're out here to help you as well.' Our phone number is all over the place with Twitter because every time we respond, our help center phone number is in that tweet."

Consider Twitter an Early Warning Signal

"We've become the canary in the coalmine to some degree with this system," said Dover. "The students will be posting out there, and we'll start to notice volume increasing. We expect to see two, three, four [complaints] a day with a university of this size. But when we start to see five or six an hour and it starts to pick up, then we look for issues developing someplace." Twitter notifications become the help center's tip off that bigger problems are surfacing. "Monitoring what students are saying can give you a head's up even sooner than some of the monitoring tools attached to the infrastructure."

A couple of years ago, when Testement brought up posts on social media about wireless issues, she was initially spurned by U Georgia's support personnel. "They would be dismissive," she recounted. "Oh, it's just somebody complaining on Twitter, and they're not submitting a help desk ticket." From a technical standpoint, if someone didn't submit a ticket or call the help desk or use one of the other "proper channels," then it was as if the problem didn't exist.

It took "quite a bit of detective work" for EITS to figure out that the complaints were originating specifically in the residence halls — usually later at night, after the students had come back from the dining hall and settled into their rooms or dorm study halls.

Automate Alerts as Much as Possible

There are a lot of ASUs in the United States — Adams State, Albany State, Alabama State and Appalachian State, just to name a few. A first step in Arizona State's Twitter response strategy is to make sure the help center is talking to one of its own students. The university's technology office business intelligence (BI) organization helped by developing a process that searches for "ASU" and "WiFi" in tweets and then checks location information and other clues to filter out people who attend other schools. (Dover's group also performs a "spot check once or twice a day to make sure we didn't miss anything.")

From there, a bot posts a copy of the tweet and its link onto Slack, which is used for internal chat communication. Dover and a few other members of the help center receive notifications on their phones and computers when something from the bot shows up, and they'll jump on Twitter to send a quick response: "Hey, sorry. We apologize that you're having issues with the ASU WiFi. Please call us at the following number so we can help you out."

Most of the time, depending on network traffic, that entire process is "close to live," Dover said. If the student calls the help center, the support person will ask questions regarding device type, location and whether the problem is new or ongoing. Sometimes, he noted, students will attach to the wireless network as a guest, which restricts them from getting to the systems they want to access. Other times, "They're letting us know there are significant issues in an area so we can get those escalated."

Use Twitter to Spur Additional Study

Once U Georgia determined that the complaints were loudest in residence halls, EITS began gathering data in other ways — through help tickets and web forms — to measure the extent of the problem. Eventually, EITS held forums in each of the residence halls to hear directly from the students. Those sessions generated an earful and helped to formulate a roadmap of priorities.

The school also put together a mobile-only survey with a handful of questions, which was promoted by university housing as well as resident assistants and residence life coordinators and pushed through social media and e-mails to students living in campus housing. That survey, which focused solely on experiences with wireless, had a 35 percent completion rate and included an open-ended question that generated "over a thousand comments" related to getting on the wireless network and staying on, said Testement. All that work "really aligned a whole picture of what we needed to do, so we could get the proper funding for [an upgrade] project," she pointed out.

Fixing WiFi Isn't About Better Netflix Streaming

While a case could always be made that students rely on WiFi too much to power not just mobile devices and laptops but also smart TVs and gaming devices, the results of those open-ended responses at U Georgia suggested something else. "A key story we were hearing from students was [that] they were trying to do their academic work in the residence halls. And because they were having issues with the wireless network, that was compromising them completing their academic work," Testement asserted. "It was not just, 'Fix my WiFi so Netflix is better.' It really became about, 'We need to fix this issue so we can ensure that students are succeeding in their classes.'"

The result was an allocation of more than a million dollars committed to upgrading wireless service in residence halls over the subsequent 18 months. Russell Hall, with 950 students, was the first to receive a wireless makeover, when EITS installed nearly 300 access points throughout the facility over spring break in 2016. A follow-up survey to those residents found considerably higher levels of satisfaction after the changes.

Tie Social Feeds Into Student Success

Arizona State's BI efforts have resulted in all kinds of "sentiment analysis" tied to social commentary, said Dover. He expects the information culled through that work to get richer with time.

Ideally, in the future, Dover anticipates that his crew will be able to fully engage with students in the "Twittersphere." That would allow them to stay in "that mode of communication that [students are] comfortable with, and we'll be able to keep the record of that particular tweet and know which agent was helping them," he said.

That interaction, in turn, would be pushed into a system that could be used for data mining tied to improving service quality. "At our volume, it's tough to be able to look at every interaction throughout a day to determine where we have opportunities for improvement," Dover explained. "With machine learning, with some of the deep data analysis, with this sentiment analysis, it can really surface those problem points faster for us with less effort, so we can focus on them."

The results could also help feed into student success, he suggested: "Okay. They were expressing major issues with WiFi quite a few times. We notice they're having issues getting assignments turned in and on time. Is this why?"

Don't Ignore What's on Social Media

Who would think that 140 characters could communicate so much? But rather than turning to Twitter as a thermometer of student sentiment, Dover advised keeping things simple. "Don't try and eat the elephant in one bite. Figure out what the pain point is and what you want to monitor."

Then develop a plan for response with the understanding "that the plan is going to change because you're going to learn as you go along." If, as Arizona State does, you come up with scripted verbiage used repeatedly in your responses, have it vetted by the communications folks.

Turning to Twitter as a lever for institutional change may seem a stretch. "It's very easy for people to dismiss negative comments as someone just complaining," acknowledged U Georgia's Testement. However, she added, a continual flurry of negative or positive comments signals something else going on "that's worth taking a deeper dive into" and learning about.