Social Data Isn't Very Representative of "Typical" Users

All advice, even good advice, should be evaluated carefully, in light of all the other things a professional knows and believes. Consider the matter of whether there is valuable insight to be gleaned from social data. It might be a rare, or brave marketer that would flatly discount such data. On the other hand, there might be dangers in emphasizing what a brand can learn from analyzing social data.


Last year, Internet users in the US created more than 500 billion influence impressions, consisting of opinions or recommendations passed from one consumer to another. But just 11 percent of online users created 80 percent of those impressions, according to Forrester Research. In other words, it is highly likely that people who post reviews and comments are different from "typical" consumers. 


By definition social data comes only from people who are posting social content, and that audience skews very young. In fact, 20-year-olds are
twice as likely as 40-year-olds to post updates on Facebook or Twitter and three times as likely to write blog posts or create YouTube videos, says Nate Elliott, Forrester Research analyst. That might be a reasonable sample population if a brand is selling products mostly to those younger demographics.


The danger is that many brands, selling to other audiences, might draw conclusions based on "unusual" customers who also are younger than a firm's target or typical buyers. 


The other problem is that online reviews tend to reflect sentiment only at the extremes: intensely favorable or unfavorable experiences with a product. But most consumers probably do not have such strong opinions about most products. 


Just 13 percent of interactive marketers use social data for brand and campaign measurement. You might conclude that marketers are making a mistake. Perhaps not. So far, social data seems not to be very representative of "typical" consumers.