Transforming Social Media Data into Predictive Analytics
Consider it the next frontier in social media marketing: While some companies fine-tune their ability to monitor social media networks, others are going one step further, fusing social media data with predictive processes.
Interest in social media, from individuals and businesses alike, is soaring. More than 900 million people worldwide check their Facebook accounts on a monthly basis, followed by Twitter, which has more than 150 million active users. In March, LinkedIn claimed it had more than 160 million users, with two new members joining the site every second.
Companies are eagerly listening to what is being said about their brands and searching for ways to take advantage of all this data. While there have been several successful cases, plucking trends out of social media chatter is still an emerging practice, notes Gartner senior research analyst Jenny Sussin. "We think overall target market penetration for social analytics is in the five to ten percent range," Sussin says. "In terms of companies doing something predictive, I'd have to say it's less than five percent."
The Challenges
As more people offer up posts and tweets about their likes and dislikes, the idea of using this data for consumer insights is becoming increasingly enticing to marketers. Picking up on this interest, software vendors offer a wide variety of sentiment and content analysis tools, as well as dashboards and reporting features for analyzing the growing stream of social media data. Companies are also investing in social analytics as a means of measuring the cost and value of their social media campaigns.
Although it is still a new area, social analytics is beginning to show up in marketing budgets. In its survey of more than 500 executives whose companies use social analytics and intelligence platforms, Hypatia Research Group found that nearly 60 percent of the respondents plan to dedicate between 1 and 2.9 percent of their annual marketing budget to social analytics next year. Slightly more than half (52 percent) will invest between 3 and 4 percent of their annual marketing budget on social analytics, and 39.5 percent will invest more than 5 percent of their budgets.
Despite growing interest, it will "take time to fully process what is feasible in regard to capturing, managing, analyzing, and…creating social intelligence in order to take action on this user-generated content," writes Leslie Ament, the report's author and vice president of research and client advisory services at Hypatia Research.
Integrating social media data about a customer with other CRM information is one of the challenges companies face, Sussin observes. "Aligning a customer's social log-in data with the company's CRM system is not a perfect craft. A lot of the work has to be done manually, and few companies have the budget or resources to do that," she says.
Privacy concerns are another issue. Although studies have shown that more consumers expect to receive relevant ads and much of the data that is placed on social networks is publicly available, customers are wary of being tracked by advertisers. One solution is to produce targeted ads based on an aggregate collection of customers' shopping behaviors and preferences. At the same time, many companies are treading carefully when using their customers' social data to personalize offers or products.
Deriving predictive marketing decisions from social analytics is also not something that can be lumped in with social media monitoring, acknowledges Zach Hofer-Shall, social intelligence analyst at Forrester Research. "Although most companies understand the importance of social media and many take the required action to monitor and learn from the data, few use social data proactively," Hofer-Shall says. "It takes a separate strategy, separate time, and often even separate people and technologies to get out ahead of social media. Most companies remain stuck reactively monitoring what happened yesterday."
MTV Responds to Viewer Howls
As Damon Burrell, MTV's vice president of consumer marketing, learned, working with a social analytics firm is one way to uncover marketing trends in social data. The network sought the help of social media analytics company Networked Insights for its new show Teen Wolf (a dramatic retelling of the 1985 movie).
The first challenge, according to Burrell, was figuring out how to get their target audience, Millennials, interested in the upcoming show. "When we first kicked off the campaign, we said Teen Wolf would probably have some core fans [of the original movie], so we're going to play it up a bit," explained Burrell in a video testimonial for Networked Insights. "As we got started on the campaign, it became obvious, based on the social data, [the campaign] was not resonating."
It turned out the decision to pitch the new show as a "remake" of the original film was confusing to viewers. "I try to keep an open mind, but I can't say the teaser trailer for MTV's Teen Wolf remake has won me over," PopWatch's Sandra Gonzalez wrote in an online review. "In fact, I find myself at a loss. The only thing [MTV's Teen Wolf] seems to have in common with the Michael J. Fox classic is that the main character is a sports-playing teenage wolf."
MTV turned to Networked Insights for help fine-tuning its marketing efforts. Based on the social conversations about Teen Wolf and other teen topics that the analytics firm tracked on Twitter and Facebook, it became apparent that Millennials were interested in the romantic relationships between the characters.
Noting that online conversations about the show spiked around the network's plans to include a gay character, for example, Networked Insights encouraged MTV to introduce the character earlier than originally planned.
Networked Insights also advised MTV to emphasize both the romantic and gruesome aspects of the show to attract both a female and male audience, maintains the analytics firm's founder and CEO Dan Neely.
"MTV had a [Teen Wolf] campaign about being in love and afraid. The moment they put that content out, we saw how it could attract two different audiences, so we advised them to split their creativity," Neely tells CRM magazine.
Teen Wolf premiered directly after the MTV Movie Awards on June 5, 2011, and attracted a total of 2.17 million viewers—a respectable number for the network's first scripted drama in several years, according to analysts. More than 3,500 tweets were recorded leading into the premiere; that grew to almost 24,000 between midnight and noon the following day. The show continued to pull in around 1.7 million viewers a week to become the network's highest-rated new series of the season. The show has since been picked up for a third season, which will be released next year.
Burrell attributes a large portion of the show's success to Networked Insights' help, noting Teen Wolf "was probably the best socially engaged–type program that we had put together because we were using social data and social insights to help drive marketing decisions."
The use of social media data as a marketing resource will continue to grow, he adds. "We're just at the tip of the iceberg," Burrell observes. "People are looking to leverage social data for marketing…it's going to be more about how can social data move business."
@WalmartLabs Purrs with Shopycat
Retailers such as Wal-Mart Stores are also applying predictive processes to their social data. After acquiring Kosmix, a company that provides clients with social media data and analysis, for approximately $300 million last year, Wal-Mart created @WalmartLabs, a unit that combs social media conversations for retail trends.
Tracy Chu, director of product management at @WalmartLabs, laid out her department's purpose in a blog post: "@WalmartLabs is working on ways to help Wal-Mart interpret social media to predict trends and learn more about what our customers want. We are mining social media sources like Twitter and Facebook, blogs, search activity, as well as transaction data to find useful insights."
A Wal-Mart spokesperson declined to comment on @WalmartLabs' progress. According to the analytic department's blog, employees have been tracking social media chatter around holiday toys, Angry Birds merchandise, and mobile commerce. The unit also launched a Facebook app called Shopycat. The app makes gift recommendations based on an analysis of your friends' Facebook activities, including their likes, shares, and posts. Shopycat will then match your friends' interests to a giant catalog that naturally includes products from Wal-Mart, as well as Barnes & Noble, RedEnvelope, ThinkGeek, Hot Topic, and other sites. Shopycat will also recommend gift cards and give the location of the nearest Wal-Mart store so you can pick up your gifts.
Going Straight to the Source
In addition to scouring the social Web for emerging trends, companies are also making bets on new product lines based on customer feedback from social networks. Meesh & Mia, an Idaho-based start-up that is licensed to sell university-branded clothing items, regularly looks to its Facebook followers for input on designs. "What better way to find out what your customers want than to ask them?" says Kat Crozier, Meesh & Mia's senior marketing manager.
Customers on Meesh & Mia's Facebook or Twitter pages will find requests for feedback on product categories so the company can learn, for example what branded items they want to see more of. "We've asked our followers to let us know what items they have in their closet [that] they wish they could wear to a game, and we then build design concepts around those suggestions," Crozier says. "We've even asked customers to select which logo they prefer in a side-by-side comparison, and then produced the logo that was preferred by our fans."
Posts asking customers to weigh in on a future product frequently receive seven times more comments than other posts, according to Crozier. Although it is difficult to connect the company's social media strategy to overall sales, it has increased the quality of the sale, she maintains. "To be able to look at a sale and know it was Sally from Oklahoma, whose daughter just got accepted into Oklahoma State, builds a much deeper relationship with our customers," Crozier says.
Avoid These Pitfalls
Many marketers agree that social media provides a useful platform for collecting customer sentiment and feedback on a mass scale in real time. However, there are several pitfalls to keep in mind.
"Taking customers' perceptions into consideration and asking for their input…will produce more loyalty and, in turn, more sales," maintains Peter Friedman, chairman and CEO of LiveWorld, a company that helps corporations manage their social networks. At the same time, companies should avoid relying on algorithm-based models as an indicator of their customers' sentiments and preferences, he warns.
"While software analytics are useful…they just don't carry through the context and nuances of how people feel and the dynamic in a conversation," Friedman says. "Social media is a dialogue and relationship venue. The deepest value comes from actually having conversations with your customers and reviewing the content and passion in their words."
Companies should also keep in mind that comments on social media networks represent the perceptions of only one portion of a company's customer base. Marketers must remember "that their social following is usually not a completely diverse audience group," advises Scott Thaler, executive vice president and chief interaction officer for Zimmerman Advertising. "They need to look carefully at the audience mix and determine if it is the group they are developing the product to attract. It is also important to know that it is not always fans who are following you."
Remember that social media analytics is only one tool, says Lisa Joy Rosner, chief marketing officer of social analytics firm NetBase. "We as marketers spent…years creating different methodologies. It's not that social media is replacing them—it's augmenting them. "It's another data point, but it's not the data point."
The largest pitfall is assuming that social data is a surrogate for qualitative, primary research, agrees Christine Crandell, president of New Business Strategies, a B2B market strategy consultancy. "Social data provides insights into engagement and enablement of audience types and the effectiveness of marketing campaigns and programs," she notes. "It is an additional, powerful tool in the marketing tool belt, not a cure-all."
5 Social Media Trends
Which Country Tweets the Most
A 2012 study by Oxford University found that of 4.5 million Twitter posts, the United States generated 30 percent of the world's tweets, making it the leader, followed by Brazil with 22 percent. Rounding out the top five are the United Kingdom and Indonesia with 6 percent, and Mexico with less than 2 percent.
The Best Time to Approach Us
Researchers from Cornell University found that for the most part, people are grumpy when they first wake up and tend to perk up by breakfast time. Our moods start to dip again during the afternoon, only to rise around 6 p.m. (quitting time). The researchers arrived at these results by weighing positive versus negative tweets posted by more than 2 million people around the world. Plan tweets accordingly.
How We Spend Our Time Online
American Internet users now devote more time to Facebook than any other Web site, spending a total of 53.5 billion minutes a month on the world's largest social networking site, according to Nielsen's "The Social Media Report."
Which Foods We'll Tweet About
The number one tweeted food topic last year was the McLobster, according to Twitter's annual Year in Review poll. Second was Fried Kool-Aid, with Starbucks' Trenta claiming third place.
Who's Addicted to Facebook, Twitter
Women are significantly more likely to engage with social media than men, according to Nielsen's "The Social Media Report." The report states that women are 6 percent more likely to visit a social networking site during a TV program, while men are 20 percent more likely to check a sports score.
Associate Editor Judith Aquino can be reached at jaquino@infotoday.com.