Valuable Data: More than a billion opinions posted online about products & services last year
Posted by: Micah Pratt, social networking research & development
Today, more and more marketers are turning to online marketing programs because of the noticeable role that peers play in consumers’ purchasing desicions. What many of these interactive marketers are failing to realize is the volume of social data created by the influence of peers in purchase decisions. Here is an article by Nate Elliott, vice president and principal analyst at Forrester Research, about how to effective utilize social data.
HOW TO: Utilize Social Data More Effectively
Last year, American consumers posted more than a billion opinions about products and services online, according to data collected by my company. With peer influence playing such a prominent role in consumers’ purchase decisions, it’s no surprise many interactive marketers are tapping into that influence via viral marketing programs and influencer outreach.
However, the vast majority of marketers ignore the staggering volume of social data all this influence creates. And those who do study social data typically use it for the wrong reason: to measure the brand impact of their marketing campaigns.
Social Data Is Often Based On Small Sample Sizes
With so many opinions posted online, you might assume it would be easy to find a reliable sample of data to analyze for brand impact. But even popular consumer brands often find it difficult to collect usable social data.
For instance, I recently reviewed a listening report for a global sporting goods brand — one that sponsors leading teams and athletes around the world and has strong brand awareness. I was surprised to see that the brand was mentioned in social media only a few hundred times each week in the U.S., and less than 100 times each week in other key markets around the world.
To make matters worse, the low quality of many social sentiment analysis tools reduces sample sizes further. When listening tools can’t decide whether comments are positive or negative, they’re usually labeled as having “no sentiment.” Three-quarters of the mentions for this brand were tagged as such, leaving less than two-dozen weekly usable posts in some markets.
If you asked your market insights team or your survey provider to analyze 25 consumer survey responses, they’d tell you it’s impossible to find statistical significance in such a small sample. The same standards must be applied to social data as well.
Valerie Jennings 
Cassie Stubs
Katie Stumpf
