Social media analytics is one of the most exciting new frontiers in business intelligence (BI). As I noted in a recent blog post, it refers to the application of BI tools, such as reporting, dashboarding, visualization, search, event-driven alerting, and text mining, to information that originates as messages streaming from social media such as Twitter and Facebook. 

Forrester sees growing adoption of social media analytics across the entire customer relationship management (CRM) life cycle. This makes perfect sense, because social media are where customers spend more and more time, voice more unvarnished sentiment, and interact with a growing range of trusted commercial enterprises in addition to their friends and families.

Recognizing this trend, enterprise CRM professionals everywhere have incorporated social media into their public relations, product management, marketing, sales, and customer service processes. In addition to establishing their brands’ presence in the leading social media communities, companies have implemented tools to support continuous listening and engagement with customers, prospects, and the world at large through these channels.

Listening and engaging via social media involves much more than BI dashboards to monitor mentions on Twitter and the like. It may also require tight integration with the company’s CRM, enterprise data warehouse (EDW), business process management (BPM), business rules engine (BRE), complex event processing (CEP), predictive analytics and data mining (PA/DM), text analytics (TA), social network analysis (SNA), and other key tools and platforms. We often refer to this cluster of technologies as enablers for “social CRM.”

When decked out in full technological regalia, a social CRM capability can involve, at the very least, any of all of the following core functions: 

  • Monitoring ongoing social media discussions related to your company, brand, products, and/or marketing campaigns;
  • Measuring awareness, sentiment, propensities, and influence based on the ongoing conversations in social media;
  • Identifying influencers, customers, prospects, and others who are engaging in these social media discussions;
  • Correlating social media identities and other qualifying information with customer information from your EDW, CRM, sales force automation (SFA), and other systems;
  • Detecting unfavorable sentiments, misinformation, complaints, confusion, and other issues that, if not addressed promptly, could impact customer satisfaction, loyalty, renewals, upgrades, new sales, revenues, profitability, and growth;
  • Auto-generating and escalating alerts and workflow action items to the relevant brand, marketing, sales, and customer service personnel to resolve these issues by reaching out via social media and other relevant channels 

Though many of these functions can be automated to a considerable degree, a social CRM platform must also deliver sophisticated analytics to those human beings — especially your customer service agents — who are the critical links in the customer-engagement cycle. Even if content analytics technology were perfected and all issues that surface through social media could be identified unambiguously, there is still no substitute for humans — let's call them social CRM analysts — to evaluate the full CRM context, meaning, stakes, and response priorities.

Of course, it’s important to underline the fact that content analytics technologies, including natural language processing and sentiment analysis, are far from perfect. That poses a problem for human beings on the receiving end of social media analytic visualizations such as dashboards, reports, and streams. Not only must your engagement agents deal with the growing volumes of tweets, Facebook updates, and other social media content, they must also sift through a significant number of issues that have been incorrectly tagged by your social media analytics tools. This is the exact same problem that you and I have to deal with in sorting through our junk email. We must distinguish false positives from false negatives, ignoring the former and restoring the latter to our inbox, where it represents a task that we must follow up on.

Even if supersmart linguists were able to perfect their content analytics algorithms, it wouldn’t address the special problems that the mounting volume of social media creates for human analysts. As social media traffic volumes continue to grow beyond the limited capacity of the human mind or workday to absorb, your social CRM teams may quickly find themselves bombarded with more automated event-driven “customers are complaining” alerts than there are stars in the night sky.

Before that day comes, you must work through several business-critical issues related to social CRM. Can you afford to monitor all social media traffic? Can you afford to respond to all issues immediately, or delay some less critical issues to later? Does it make sense to respond to all issues, or simply a prioritized subset of issues associated with social media content produced by the most influential users? Should you outsource the social media listening/engagement functions in order to leverage the scalability and automation that an outsourcer can bring to this increasingly critical function?

Yes, outsourcers are human too, and face the same scalability issues that you do if you choose to manage social media engagement in-house. However, many outsourcers who have incorporated social media into their multichannel CRM offerings are evolving their social media analytics capabilities — as well as their professionalism, specialization, and best practices — to deal with this issue comprehensively.

Look at it this way. You may already be outsourcing your call center, which is a key piece of your CRM infrastructure. Doesn’t it make sense to at least consider outsourcing your social media listening and engagement functions as well, so as to avoid getting swamped by this new tsunami of a CRM channel?