by Kim S. Nash

Inside 4 Companies’ Facebook and Twitter Strategies

Nov 19, 2010
CRM SystemsE-commerce SoftwareInnovation

Capturing customer data and building communities on Facebook and Twitter isn’t easy. Learn how four CIOs are integrating social media with business data to draw more attention to their companies’ products and deliver better service.

The customer may always be right (even when he’s not). Now he’s also in your face, thanks to social media.

On Oct. 6, the clothing store Gap surprised the Web with a new corporate logo and was immediately deluged by thousands of comments—nearly all complaints—on Facebook, Twitter, various blogs and other sites.

The logo—round black letters on a grainy white background—“blows,” someone wrote. “Incredibly cheap looking.” “Hideous.” Some Gap fans took to calling each other names, like “sissy.” Others made the kind of pronouncement that causes every marketer to shudder: “I’m never shopping there again if you proceed with the logo change.”

By Oct. 11, Gap said it would drop the new logo, returning to the 20-year-old version with white Spire font on a navy box. “We recognize that we missed the opportunity to engage with the online community,” said Marka Hansen, president of Gap North America, in a statement.

Maybe you think Gap overreacted. That nobody likes change and social media by its nature encourages users to find something—anything—to talk about. Gap’s logo was simply the topic of the day. Skeptical bloggers even speculated that the logo change was a publicity stunt. “Just because there’s a big uproar on social media doesn’t mean it means anything,” says Zach Hofer-Shall, an analyst at Forrester Research.

On the other hand, an uproar might be significant, such as when users of some of Dell’s laptops complained—legitimately—about sound-quality problems. But weighing the value of data emerging from social networks is tricky. New social media monitoring tools can identify and capture online conversations that meet user-defined criteria. Yet these tools typically don’t support the deep analysis companies can do with traditional business data. Nor do social media tools easily integrate with existing CRM suites and data warehouses, which would allow easy data sharing.

Todd Michaud, vice president of IT at Focus Brands, would like to combine social media data—what people are saying online, the locations they visit—with his internal demographic and business-intelligence data to create a more complete picture of his customers and to fine-tune marketing efforts. “The real benefit of social media is to integrate it with in-house data you have in your business intelligence, CRM and customer call-center systems,” says Michaud, whose company is the franchisor of the Carvel, Cinnabon, Moe’s Southwest Grill, Schlotzsky’s, and Seattle’s Best Coffee food franchises. “But there’s a void in that space.”

While the software evolves, milking social media requires more human intervention than you may think. That can include cutting and pasting the social media comments collected by automated tools into spreadsheets for later analysis or having people sit before a bank of monitors tuned to various social media sites, scanning for relevant conversations to join.

CIOs must also get IT and marketing to work well together—far better than they do now. Innovative companies try to appeal to increasingly tech-savvy consumers through social networking and mobile applications, says Robert Urwiler, CIO of Vail Resorts. IT knows the technology; marketing knows the customer. “It is a huge opportunity missed,” Urwiler says, “if IT and marketing are at odds.”

Customer Data of Your Dreams

The amount of data a CIO could glean from social-media chatter could smother the typical corporate network. Consider that the average Facebook user is connected to 130 friends. And among Facebook’s 500 million users, half log on to the social network every day. LinkedIn boasts 80 million members, with someone new joining approximately every second. On Twitter, about 55 million tweets are sent forth daily by the network’s 106 million registered users.

That’s a lot of sharing.

Typical CRM suites, such as those from Oracle, and SugarCRM, were built to capture consumer-to-company interactions. A customer dials the call center and an agent records notes from the conversation. Or the company logs and sorts queries to its website. Data can be parsed to show a single customer’s history with the company or the activity of customer segments.

However, users of Facebook, LinkedIn, Twitter and other social media typically talk to each other; it’s consumer-to-consumer, rather than consumer-to-company, commentary. What companies seek are the sentiments about brands and products within those conversations. A recent tweet from Twitter user @BaileyKathlyn, whose bio lists her as a “foodie/cooking pro,” asked her followers about what gets them through a Monday. “I rely on Starbucks French roast, Sirius NPR station and Angry Birds Halloween edition,” she wrote. Coffee, radio and video games—you can almost picture her at her desk. If aggregated, sifted and analyzed, such posts can provide potentially valuable clues about consumer behavior.

But when social media users name products or brands, companies monitoring those mentions have no easy way to know if they are real customers. If a company could cross-check, say, Twitter handles with names in its CRM system, then analyze Twitter content against CRM data, the results could be the most revealing—and eventually lucrative—development in Web customer service to date, says Henning Hagen, a principal at Booz and Co.

Companies first built websites to draw people, Hagen observes. Then they sought to turn that traffic into cash by selling products, selling ad space or harvesting sales leads. Now, Hagen says, the next stage is upon us: Companies can and should use social media to form deep, personal relationships with customers. “There’s the potential to be much smarter about what your customers are doing and when, and finding ways to keep in the forefront of their minds,” he says.

The way to do that, however, is not simply to collect any mention of your brand uttered online. There’s a difference between loving the beach and taking home every grain of sand. Distinguishing useful data from the noise created by all those millions of social users is what’s important, says Michaud at Focus Brands.

If a Twitter user tweets about your product, it means little if you don’t know who he is and what else he says online. But if you can discover that information and overlay it with your company’s other data about him, Michaud says, business analysts can start to predict what he’s likely to do next, salesmen can devise offers for a product he can’t resist and marketers can tweet him about it directly. “Customer service interactions, business we’ve done together and what you’re saying in the online universe help give me a picture of you,” he says.

The problem is that the tools needed to make that lovely corporate dream come true don’t exist. Not yet, anyway. Tools from companies such as Visible Technologies and Lithium can comb the Web looking for mentions of your company’s name or key products, says Hagen. But they can’t automatically restructure that data in a way that would let analytics packages decipher it.

At many companies, therefore, interpreting social media is a time-consuming job done by humans. At Focus Brands, some analysts following online conversations cut and paste information from social networks into databases. “But we don’t want to throw 30 people into managing Twitter,” Michaud says.

His IT team is building features that would let the company’s salesmen link to the Twitter streams of their top franchisee prospects. Next, he wants to attach social media information to data about that franchisee in a corporate CRM system. The company currently uses Microsoft CRM for customer service and for its franchisees. “The vision I have is, when you drill down on a record, it will show the person’s LinkedIn profile, Twitter stream, Facebook page—whatever I can get that’s public.”

Early Warning System

Many companies may not be ready to—or may not want to—collect, dissect and fully absorb social media into enterprise systems. Instead, they monitor online conversations to get a jump on potential problems.

But they have to pay close attention, especially to Twitter, where millions of posts fly by every day. The average tweet is pertinent only for about an hour, according to Sysomos. The company’s recent analysis of 1.2 billion tweets found that just 6 percent were retweeted by others and 92 percent of those retweets occurred within an hour of posting. Less than 1 percent of tweets are repeated by the time they reach the ripe old age of three hours.

To observe the ebb and flow of online material relevant to it, computer maker Dell built a command center where servers run tools from Radian6 that pick out posts on Twitter, Facebook, LinkedIn, and various blogs and websites that contain keywords such as “Dell” and the names of its products. Customer service agents watch for complaints or problems they can address in real-time. More generally, sales watches for leads, such as individuals posting unhappy comments about competitors’ products. Human resources systematically monitors LinkedIn for potential new hires.

While some companies want to ban the use of social media at work, Dell promotes it. “We democratized and have given the tools to many organizations,” says Manish Mehta, vice president of social media and community at Dell. For example, tech support, sales and marketing staff have access to information from the social media command center, as do human resources managers looking to recruit. Mehta is also installing command-center monitors in the hallways outside Dell’s senior executive offices. “That’s how deeply we’re embracing social media.”

If, for instance, an exchange with a customer on Twitter or Facebook flowers into a full-blown tech-support case, the Dell employee copies and pastes the pertinent data—name, contact information, product, problem—into the company’s case-management system, Mehta says.

Mehta wishes this task were automated, but he expects software vendors to develop such tools soon, as they did to help simplify companies’ move to the Web 15 years ago. “Within a couple of years, hundreds of companies surfaced to fill that gap. Same thing will happen here.”

In the meantime, chatter picked up by the command center warns Dell about potential widespread problems with its products. After shipping one type of notebook recently, Mehta’s team started to see complaints on Twitter and blogs about poor sound quality. Yet Dell’s call center hadn’t received such complaints and quality-assurance tests had shown the sound was good, he says.

After a few days, Mehta brought verbatim comments to product engineers who, with those clues, traced the problem to a subwoofer that was inadequate for rich low notes. Dell changed subwoofers in subsequent machines and the online complaints disappeared. “We averted a larger problem,” Mehta adds.

Creating Community

Vail Resorts, which runs six ski resorts, including Vail and Beaver Creek, kicked off the ski season last month with a social media push aimed at building its already high-rated brand.

The strategy is to extend the ski culture—which is naturally prone to exuberant sharing—to the online realm in a way that showcases the Vail resorts, says Urwiler, the company’s CIO. “This is not at all about tracking people. It’s more about giving customers a voice to talk to more and more people about their experience at Vail.”

Rather than market itself directly through social media, Vail Resorts uses Visible Technologies tools to find posts on Twitter, Facebook and other sites where people comment about Vail’s properties. Then marketers spotlight what customers say online about the resort—and what they do on its mountains. Skiers like to broadcast this information in status updates and shared photos, says Mike Slone, Vail Resorts’ director of interactive marketing. “We can talk all day about ourselves on Facebook and Twitter. But if we have thousands of people on a daily basis sharing about us, that’s more authentic than anything we could do on our own.”

Vail Resorts is trying to make it even easier for skiers to share. Two years ago, the company began to embed RFID chips in ski passes to speed up lift lines. Now the chips can track such data as the amount of time a skier spends on the slopes, where he or she has been, and the vertical feet covered in each run. A newly deployed Web and mobile application called EpicMix can automatically turn that data into Facebook or Twitter posts for customers who sign up for an EpicMix account and agree to share their information. Vail marketers might also retweet or share those updates.

Meanwhile, Vail has EpicMix users in its customer database, which already has information about season pass holders and other regular visitors, says Urwiler. For now, though, he is not concerned about integrating social media data with Vail’s CRM system. Knowing who the customers are matters less at the moment than making sure they’re talking about their experiences.

Vail does, however, offer customers points toward unique collectible pins and other rewards as they add to their stats via EpicMix. And the company looks for opportunities to engage directly with potential customers through social media. A handful of marketing staff, Slone included, collectively spend about 40 hours per week using the Visible Technologies tools to identify people using keywords, such as “skiing” and “snow.”

Last year, Slone spotted a Twitter post by a man who planned to take his family to Breckenridge, Colo., where Vail operates a resort. Slone jumped. “It turned into a 16-post conversation where we helped the guy choose lodging, dining and a sleigh ride,” Slone recalls. “There were e-mails and phone calls in there, too.” Vail beat out a competitor who also tried to engage the vacationer on Twitter; plus, the customer tweeted good things about Vail a couple of times.

The personal touch matters. No software could have participated as well as a Vail employee in that exchange, Slone says. “We’re all humans. We want to engage in conversations with other people, not robots,” he says.

Blending IT and Marketing

Vail is one place where marketing and IT work hand-in-glove, Urwiler says. Marketing, IT and Vail’s CEO together devised the company’s social media strategy. Marketing vetted the tools, and IT installed and supports them.

But that’s not common, and social media projects will quickly expose that disconnect. When CIOs and chief marketing officers don’t collaborate, departments may undertake competing social media projects. Or groups may bring in software IT hasn’t approved, potentially causing support and integration problems, observes Focus Brands’ Michaud.

In a study by Accenture and the CMO Council, a skimpy 4 percent of marketing executives and 7 percent of IT leaders polled said their companies are very prepared to exploit digital marketing channels. (See “CMOs and CIOs: Can This Relationship Be Saved?”) Yet as unprepared as they think they are, marketing and IT seem to be battling to lead their company’s digital marketing strategies: 58 percent of the IT respondents surveyed said they were in charge and 69 percent of the marketers thought they were.

Well-suited to lead this work are IT leaders with a penchant for marketing, says Hofer-Shall from Forrester: Managers who can imagine the business benefits of using social media and who can fill in the technology pieces as projects progress.

Greg Flesher combines such skills as a vice president of academic e-learning at Elsevier’s Nursing and Health Professions division, which publishes academic materials for students in technical or professional education programs. The products include textbooks, journals and teaching materials. Elsevier is a unit of the $9.7 billion publisher Reed Elsevier.

To increase word of mouth about these materials and, he hopes, revenue, Flesher’s group is orchestrating a social media project focused on the nursing market and working with IT, sales and marketing.

Through nursing conferences and training programs, and by observing collegiate social media activity, the business unit has gathered a group of health-science students to become Elsevier Student Ambassadors. On Elsevier’s websites and its pages on Facebook, Twitter and YouTube, the students share links to relevant healthcare material from around the Web. The links generate traffic for Elsevier and heighten awareness of its brand. The ambassadors are paid in both money and textbook credits.

“We’re trying to create interaction where the students are—saying you don’t have to come to our site for everything,” Flesher explains. “But the link-through is important.” Links that lead more visitors to Elsevier’s website and learning materials help build brand awareness, he says, and can lead to more sales.

Flesher’s background is equal parts technology and business: He has built e-learning and testing products that Elsevier sells, such as online courses and learning-management systems. In a previous position, as associate director of emerging technologies for the company, he researched and tested new software. He chooses for his staff employees with business and technology experience similar to his. “[We] sit in the intersection between sales, marketing, IT and production.”

Because, at various times, he’s reported either directly or through a dotted line to different facets of IT and marketing, Flesher has formed relationships across departments. That’s helped him, he says, in his relatively new role establishing an e-education department at Elsevier. He and his staff are “able to fill IT knowledge gaps on the fringes of other departments.”

For Flesher, the next challenge is to aggregate and analyze the social activity against known information about customers in Elsevier’s system and its homegrown CRM application. The company is working with Parature to build interfaces between the systems. Parature consultants created a report that shows posts and comments containing certain keywords and specific technical-support requests. Ultimately, these will be combined with CRM records, Flesher says. For now, however, the information is stored in Excel spreadsheets or PDF documents for product developers working on new versions of Elsevier’s teaching materials.

Elsevier is now researching the idea of “personal learning networks.” Students and teachers could use Elsevier-run communities on Facebook to discuss coursework that is linked to the Elsevier website. “It’s important to allow students to go beyond the boundary of classroom,” he says. “Publishers have to adapt.”

What You’ve Always Wanted

Companies know social media has changed the rules for interacting with customers, and astute companies are looking for ways to figure out what it all means. Given the way events such as a new-product launch or a corporate acquisition generate spikes in social media traffic, “it will take more advanced integration with CRM to test whether sudden uproars are worth responding to, whether they mean anything for your particular customers,” says Forrester’s Hofer-Shall.

Though social media monitoring tools lack the power of enterprise analytics, there are plenty of consultants willing to do that work, he says. Or you might, like Vail and Dell, prefer to understand how your customers interact with social media firsthand.

In some ways, Hofer-Shall says, information gleaned through social media is easier to understand than other kinds of customer data. E-commerce software may log how many times a customer places four items in her shopping basket but buys just one. And CRM analytics can triangulate those facts with other data to suggest possible explanations. But those conclusions can still be wrong. With social media, Hofer-Shall says, “you know when someone’s saying something positive or negative.”

On Twitter one recent night, @aebthestyleicon (an underrated style icon, according to her bio) posted, apparently happily, “Watchin Entourage and eatin a Cinnabon :-)!!” It’s information like this, says Michaud, that will help his company create a 360-degree view of that customer.

“For ages,” Hofer-Shall adds, “we’ve been trying to get at this kind of [behavioral] information from customers. Now we have it unprompted.”

What you do with it will be your business.

Follow Senior Editor Kim S. Nash on Twitter: @knash99.