Companies have shied from sifting through social media sites because they worry about not seeing a return on investment and don't have the necessary resources to invest in tools that analyze customer data.
That finding came out of a recent survey from The Data Warehousing Institute (TDWI), which queried more than 450 business executives and IT and data professionals about their companies' social media analytics practices.
The survey found that less than a third of the respondents effectively use social media to help manage their brands. These companies have minimal involvement with social channels because they lack the budget and staffing to use proper customer analytics tools, the study said. They also fear they won't see a return on investment (ROI).
"They can't define a hard return. And not only do the CEOs want to see that, but [so do] the CIOs," according to David Stodder, TDWI's director of research for business intelligence (BI).
Additionally, tension exists between company departments over the sharing of social data, said Stodder, who authored the study. "Social media analytics is becoming a new silo," he said.Social data has created a new tug of war between departments that are ready to use the data quickly (such as sales and marketing) and IT, which insists on clean data that meets company standards, he said.
But without fully knowing customers, businesses lose the chance of identifying the correct marketing segment for them, Stodder said. When companies later attempt to extend loyalty programs, a lack of knowledge prevents them from understanding when those perks should be offered in a customer's life cycle.
Lacking this insight, marketing managers can't determine when customers will pursue another brand, what products customers should have been offered to prevent them from breaking away, and when the offers should have been made, the study said. Social media analytics answers those questions and helps businesses improve marketing by quantifying customer value at different stages of the life cycle, the study revealed.
"Customer loyalty sounds important, but it won't work if you're not addressing customers where it's going to be profitable," Stodder said.
The TDWI study found that most organizations haven't invested in tools to monitor social media sites; 65% use OLAP and BI. According to the study, these older technologies are used even by companies that are proficient with other methods of analyzing social media, and they continue to do so probably because they are have existed longer than social media and are ubiquitous in business.
To stay current, companies will need to pursue in-memory computing to provide rapid analysis of large volumes of customer behavior data, the study said.
For more on social media
Learn how social media revenue is expected to soar
Businesses should also consider other cutting-edge tools and methods for data mining, advanced statistical analysis and predictive analysis, the study said. These tools and methods enable businesses to examine large data sets -- such as transactions, loyalty card memberships, clickstreams and other sources -- more rapidly to test hypotheses and forecast buying behavior, the study claimed.
The TDWI study revealed that more than 30% of respondents have started to use advanced analytical tools. That figure may seem low, but it is trending upward, Stodder said. If such questions were asked two years ago, only about 20% of companies would have gotten that far, he said.
Still, the majority of companies outsource their social data analytics. "There's still a reaching out to specialists. There's still a lot of room to grow," he said. Outsourcing happens, in some part, because vendors have yet to offer a product that fully integrates the various social tools within the enterprise, Stodder said.
More than a quarter of the study's respondents revealed that gathering and analyzing data in real time was one of their more challenging issues. The faster organizations can analyze customer data, the sooner they can set their marketing and sales processes in motion to react to customer demand, Stodder said. However, social media intelligence systems can't function within enterprise data warehouse architectures that take too long to extract, transform and load data, he said.
According to the study, master data management (MDM) can help businesses simplify data integration by establishing a reference source or registry that provides a common set of business-oriented customer definitions. Businesses that have established MDM are able to use this higher-level reference to connect business names or objects in multiple data sources. But MDM apparently hasn't taken hold yet in analytics: only 27% of those surveyed by TDWI have implemented master data management for customer data.
"It's amazing. There is a lot of new data moving in, and it's tough to follow with customers changing their minds," specifically about companies and brands, Stodder said.
Not to mention, businesses have to overcome challenges to access data from the two largest social media sites, Facebook and Twitter, Stodder's study revealed. The "big pipe" of data from the two websites is no longer openly available to social media analysts. Facebook has controls built into its API to guard user privacy, and analysts have to depend on seeing only user "likes" on public pages. Complete access to Twitter's data hose of more than 350 million tweets a day is available only through the website's partners, Gnip and DataSift.
Not knowing how customers will respond to products is tough for businesses to monitor, Stodder said. But he added that this new field is also "cool" for all the marketing and sales possibilities that didn't exist prior to social media.