What do corporate heavy hitters have to say about The Future of Social Media in 2010? A recent Silicon Valley American Marketing Association forum explored innovative ideas and best practices for leveraging social media, with insights from the likes of Ed Terpening of Wells Fargo, Jeanette Gibson of Cisco, and Maria Pomeromo of Adobe.
Adobe, Cisco, and Wells Fargo have all made significant commitments from the top of their organizations to better engage with customers through social media. However, many companies are not gaining the maximum impact from their customer listening –and that got our data analytics team thinking…
6 key insights on using text mining in social media
1. Text mining can make social media efforts more effective
If your company is still taking a manual approach to analyze customer feedback in social networks –you may be missing something vital. As the volume of comments from blogs, Twitter, Facebook, LinkedIn and other sources continues to grow, manual analysis can’t keep up. Marketing and product management teams risk missing valuable insights, and this is where text mining comes in.
2. Let the data talk to you
By systematically categorizing comments, applying a metric of importance or severity, and tracking trends over time, you can uncover what customers think about your products and brand. Browsing blog posts and comments are no longer enough, and in fact, present a big risk: you may focus too much on what you’re looking for and too little on what the majority of current and potential customers are really saying.
3. Take an enterprise approach to evaluating customer feedback
Instead of monitoring online conversations only at the product level, think about expanding to a broader, integrated view of what customers are saying. They see your company as one entity, even if lines of business or products are managed semi-autonomously. If you only monitor dialog around a few key products, you may miss out on potential insights by not putting all of the pieces together.
4. Text analysis: the best thing since sliced bread?
An automated text tool can help you zero in on key themes, and the associations among them. For example, if you look at blogs about Toyota, there has always been a customer conversation about quality; however, lately, there’s been an increase in the total volume of comments, and the issue of Toyota quality is being discussed in more negative terms. Using text analysis, and social network data from Twitter, Facebook and other sites, you can measure and track sentiment and its intensity over time. This can provide a kind of gut-check, or early warning system, well before you tally up your next Net Promoter Score.
5. Bionic ears, on Steroids
Automating the analysis of listening efforts lets companies integrate analytics into their social media monitoring. Companies like Adobe are already using brand monitoring tools to enhance their customer listening capabilities. Leaders in this space include Radian6, Cymfony, and SM2 by Alterian, and while these solutions currently have limited analytic capabilities, you can leverage the content these tools provide.
Radian6, for instance, allows you to monitor and export data from blogs, top video sharing and social networking sites, forums, and mainstream media sites. Suck it into text mining software like SPSS Modeler, and you’ll hear things from your customers that never used to hit your radar. Ready to take it up a notch? Add this layer of information to other feedback sources you track, including customer service email, call center notes, and customer surveys.
6. Your customers are talking behind your back
Turn around, already. You can use these techniques to create a comprehensive Voice of the Customer program that gives you a rich, nuanced view of what’s on customers’ minds. Companies like Adobe, Wells Fargo, and Cisco have done a great job of diving into massive amounts of customer data. Their next step—and the future of customer listening—is to mine it for the real gold.