4 Ways to Actually Use Big Data
Just a few days into 2014, Southwest Airlines, the biggest discount carrier in the U.S., was put to the test. Wintry weather in the Midwest and Northeast had forced the sudden cancellation of hundreds of flights. Thousands of stuck passengers reached out to phone and web support to rebook, only to find the systems overwhelmed. So, they turned to the last resort of the stranded traveler: Twitter.
12,000 tweets--many of the extremely irate variety--flooded into Southwest over the course of six stormy days, with totals spiking as high as 3,000 tweets per day. On the other end, as the airline’s last line of defense, stood a staff of just seven social-media experts, according to a USA Today article.
While this may be an extreme case, Southwest is hardly alone in being swamped by social media demands and data. Large consumer brands like Coca-Cola and the NBA register tens of thousands of social media interactions every day, from mentions on Twitter and "Likes" on Facebook to customers posting questions and complaints and bloggers sharing the latest corporate news.
Slowly, however, some companies are finding ways to turn this data headache into a resource. Big Data--the name given to the deluge of digital interactions companies have with clients--has been hyped for years as a kind of “crude oil” of the new millennium, hugely valuable but useless if unrefined. The challenge is that advanced software suites and analytics experts have generally been needed to make any sense of the terabytes of raw information that can be collected daily.
Social media, however, is increasingly offering ordinary companies entree to the esoteric world of Big Data. Armed with new, user-friendly analytics tools, businesses are finally finding practical--and cost-friendly--ways to process the gusher of digital information on clients and prospects. Intuitive, subscription-based services from upstart analytics players like Brandwatch, uberVU, HootSuite [my company], and others have started to democratize the kinds of insights once available only to the tech-savvy few.
In the case of Southwest, their social media team was not only able to weather the big storm but emerge from the episode riding a wave of positive sentiment and with more followers than ever. Here’s guessing Big Data--and the right tools to crunch it--played a big part in that success. For businesses looking to leverage social data to their own advantage, applying these basic analytics hacks can be a good place to start:
Use filters to separate important messages from background noise: Computer company Dell famously receives more than 25,000 mentions daily on social media channels in 11 different languages. The overwhelming majority of those--from a practical business perspective--are low priority. But by using social media analytics tools, Dell is able to automatically filter for the messages that actually matter: the ones from influential Twitter users with thousands of followers, the stories posted on respected tech blogs and forums, the urgent customer requests that could go viral if left unaddressed.
These tools each apply their own proprietary algorithms to identify the most pressing messages in real-time, taking into account keywords, views, and other customizable fields. The end result is that a flood of social media data is reduced to a manageable stream. Companies can then initiate immediate triage in the event of bad news, help spread good news, and relay concise information to team members in marketing, sales, customer service, or other departments for follow-up.
Track overall changes in message volume: On the evening of September 2, 2013, the number of Tweets about British Airways began to spike abnormally. And the message wasn’t good. Disgruntled Twitter user Hasan Syed, angry over bags the airline lost, fired off a message that read, "Don't fly @BritishAirways. Their customer service is horrendous." Only instead of sending it off to his followers, he paid to have the message tweeted to an estimated 50,000 other users in New York and the United Kingdom, two of the airline’s key markets.
Incredibly, it was a full 10 hours before British Airways staff finally acknowledged the message and, belatedly, attempted to put out the fire with an apology. By that time, the story had jumped from social media and spread to news outlets on both sides of the Atlantic, from the BBC to Time.
A lot of trouble could have been saved in this case with some fairly simple analytics tools. Changes in message volume on social media, like the one British Airways experienced, often indicate that something significant is up--either good or, as in this case, bad. Alerts can be set up to monitor mentions of a company’s name and other keywords and issue warning emails when any unusual spikes in activity occur. This enables businesses to get ahead of P.R. mishaps, address concerns, and prevent incidents from spiralling out of control.
Incorporate tools that automatically track sentiment: Back in 2009, one little booger caused one huge problem for Domino’s pizza. An employee filmed himself picking his nose at work, then uploaded the video to YouTube. Predictably, the footage went viral, seen at least a million times before being pulled down. For Domino’s, the video and ensuing uproar--outlets from The New York Times to NBC ran stories on “booger-gate”--sent already low customer ratings into a downward spiral.
To fight back, Domino’s changed their recipe, offered money-back guarantees, and set up a microsite where diners could upload real shots of their food. Meanwhile, they tracked changes in public opinion on social media to fine tune each aspect of their campaign. In the end, this multi-million-dollar “mea culpa” was a dramatic success. Sales in U.S. locations increased 14 percent in the quarter after the campaign and share price rocketed 75 percent the following year.
Today, this kind of nuanced sentiment tracking can be accomplished instantly. Social analytics software--which automatically scans the text of thousands of messages to reveal share of positive, negative, and neutral sentiment--can give companies a real-time window into how consumers feel about their product, their brand, competitors, or any combination of keywords. By monitoring search terms overtime, brands can see how opinion evolves in response to events inside and outside the company and shift strategy accordingly.
Choose software that spits out snazzy reports: Wintry weather isn’t the only PR challenge Southwest Airlines has faced in recent years. There was the time a pilot accidentally left his radio on, filling Texas airspace with a homophobic rant about his flight attendants--an incident that drew more than 10,000 social media mentions. And then there was the time a hole opened up in a fuselage mid-flight; passengers taking advantage of free Wi-Fi tweeted about the incident as it happened.
Yet, through it all, Southwest, has managed to avoid major wounds and, in fact, was named number one in customer service in the 2013 Airline Quality Ratings. A big part of that, no doubt, has to do with the carrier’s passionate social media community, which now numbers 1.6 million Twitter followers and 4.2 million Facebook "Likes." Southwest’s Social Business and Listening Team has spent years cultivating and expanding this community, improving the airline’s visibility, boosting loyalty, and helping it to grow despite the occasional P.R. setback.
Still, communicating the value of social media to company executives focused relentlessly on the bottom line isn’t always easy. Deep skepticism persists among many executives, for whom Twitter and Facebook represent little more than funny cat videos or--at best--“soft” tools for networking.
Analytics offers an antidote. Current tools often include sophisticated reporting functions capable of spitting out boardroom-ready charts tracking changes in brands’ overall visibility and sentiment with time or comparing growth against competitors. Armed with these visual aids, marketing and community teams can show stats-minded executives that social media actually makes sense for business.
RYAN HOLMES | Columnist
Ryan Holmes is the CEO of HootSuite, a social media management system with eight million users, including 79 of the Fortune 100 companies. In the trenches everyday with Facebook, Twitter, Google+ and the world's largest social networks, Holmes has a unique view on the intersection of social media and business. Follow him on LinkedIn.