Social Listening Heads Inside the Contact CenterSocial Listening Heads Inside the Contact Center
A data scientist shares how his firm helped Motorola better understand online conversations for customer support and other key business processes.
October 31, 2014
A data scientist shares how his firm helped Motorola better understand online conversations for customer support and other key business processes.
Mention to a friend or colleague that you've had a bad customer experience or some difficulty dealing with a company, and chances are you'll be asked -- maybe even admonished -- "Did you tweet about it?"
Whether you're dealing with lost luggage, shoddy merchandise, or a lame return policy, venting on Twitter and other social networks is quite often the thing to do these days. Searching the social Web for what others have to say about the same complaint is a natural next step (think #iphonefail) -- and then you pick up the phone or head to the company's Website for customer service. Pity the agent who has to handle your inquiry clueless to all that's already been said about a product or a service in social network conversation.
As an IT professional, you might be tempted to dismiss the need for social listening integration with an, "Oh, marketing can deal with that." The truth is, you've got to get on top of the issue and figure out how best to give contact center agents the social context they can use to optimize the customer experience at that touch point. Helping you do so is certainly among the goals IBM and Twitter spelled out earlier this week when they announced they'll be integrating Twitter data into IBM's cloud-based analytics, customer engagement platforms, and consulting services.
Despite this week's chatter around the IBM and Twitter teaming, feeding social insight into the contact center is something any number of organizations have been doing or working toward for a couple of years now. Motorola Mobility Service & Repair is a case in point.
Two years ago, Motorola partnered with a Chicago-based data consulting firm called Datascope Analytics on a social listening project. I learned about the project from Mike Stringer, Datascope co-founder and data scientist. Stringer couldn't get into the nitty-gritty, but did give me a good general overview.
Motorola approached Datascope about two years ago, knowing that a lot was being said out in the wild about its smartphones and the company in general, good and bad, and that it could use the insight from those conversations to improve its products and better serve customers. With its phones sold through carriers, Motorola had a shortage of direct customer feedback, Stringer said. "It recognized how important it is to have that kind of feedback, so it wanted to build a platform that would help it really distill what people were saying about their devices."
At the time, for example, tens of thousands of sentences littered the online world -- in consumer Websites, user forums, product review sites, social networks, and so on -- about Motorola's Razr M Android phone. To help Motorola prioritize issues based on what people were saying online, Datascope used a technique it calls shared opinion grouping, Stringer said. Using text analytics and sentiment analysis, Datascope is able to gather what's being said online about Motorola products like the Razr M and features and, importantly, whether the comments are generally positive or negative. It then delivers a summary statement through a Web-based interface that saves Motorola customer service agents, technicians, or others from having to scroll through hundreds of individual comments and deciphering the overall feeling themselves.
Again, Stringer could not reveal too much about Motorola's use but I could imagine how this insight but come into play in the contact center. Say a consumer calls in with a complaint about really low screen resolution. Via the Web interface, the agent could see that "Great contrast!" is the most representative comment about the model in question, and so would know that the issue is particular to this caller rather than a more widespread problem. The agent gains a better understanding of how to engage with the caller in this situation. Perhaps, too, Motorola could use the insight to prioritize call handling depending on how positive or negative the online sentiment is on a phone or a feature.
Of course, that's just the tip of the iceberg when it comes to bringing the customer conversation into the contact center. As Bryant Harland pointed out in his NoJitter post earlier this week, Using Analytics Without Freaking Customers Out, the real value for many organizations will be when they can match up a customer's social commentary with his or her demographics, purchase history, and so on. But that, as Bryant suggests, can quickly edge over into the creepy. And sometimes, as Stringer noted, data in aggregate is just as useful in making a difference in the customer experience.
Follow Beth Schultz and No Jitter on Twitter and Google+!
