Evan Schuman's StorefrontBacktalk

Among the most frustrating data-analytics facts in retail today is that the goldmine of customer data locked in social networks is virtually untouchable because of the way the data is structured.

Think about it: millions of customers and prospects post their innermost product thoughts on Twitter, Facebook, Youtube and the blogs of themselves and their friends. This info is free and often visible to all. But how do you take that data and match it to specific customers/prospects so that the data can be acted on? That’s the untouchable part.

Google this week (August 26) tried to move in on this space with a service called Google Realtime Search. “We’ve added a conversations view, making it easy to follow a discussion on the real-time Web. Often a single tweet sparks a larger conversation of re-tweets and other replies, but to put it together you have to click through a bunch of links and figure it out yourself,” said Google’s announcement on its blog. “With the new full conversation feature, you can browse the entire conversation in a single glance. We organize the tweets from oldest to newest and indent so you quickly see how the conversation developed.”

The service itself is nice—although on the test searches we did, it wasn’t any more useful than running search.twitter.com plus any search engine—but it does little to truly connect the dots between the content and who is posting it. Some observers have been tempted to draw the conclusion that the absence of quantifiable social media data means that there’s little retail potential out there, which is silly.

Still, the opportunity to truly try and make this data useful is still there. A retail chain would presumably start with its installed base of customers and have the engine gather as many comments from those individuals as possible, dragging the results into the customers’ CRM profiles. The same would be attempted with a list of prospects.

Some companies have already done this on Twitter, although it’s typically not automated. Employees are assigned to search for the retail or brand names and then to join the conversations and fix problems. But the very nature of that effort is going to be reactive, and personnel bandwidth makes it of very limited value. (To be clear, the value is quite high. But it can only impact an extremely small percentage of a major chain’s customers and prospects. That’s what we mean by limited value.)

But a truly automated approach that focuses on customers and prospects would be proactive—and it could potentially have an impact on far more customers, prospects and sales.