This past week, I attended Social Media Breakfast 27 in Boston; its theme was “Beyond the Echo Chamber” with an emphasis on conducting market research using social media channels.
Realizing this focus, I recalled the early days of what we now call “social media” – before Twitter and Facebook – when blogging almost exclusively ruled our thinking.
Monitoring solutions (like Cymfony- now Visible Technologies- a client where I worked at the time) focused on mining blog comments in particular to glean market insights. Having recently spent time at a market research company, I was intrigued by the prospect of a ready-made group of research subjects; people who were not only easier to acquire than the typical research panel (probably one of the most difficult and expensive aspects of research), but had already expressed their opinions. One need only find, collect and analyze them.
Of course, it’s not that easy. The slow, uneasy evolution of social media monitoring tools is testament to the real challenges behind this difficult science (If I’m not misusing that word).
At the breakfast, speakers picked apart some for the challenges and dangers, along with the opportunities, in using social media for research.
They Called the Event “Beyond the Echo Chamber” for a reason
It is vital to any market research to keep in mind the potential biases of your sample, whether they be used for qualitative research (such as focus groups) or quantitative (such as surveys). Jeffrey Henning of Affinova pointed out that one must account for the tendencies inherent of the users of any platform. Surveying Twitter users? No matter their background, you are likely selecting a group skewed toward early technology adoption- certainly, they have access to the Internet via PCs or mobile devices that many on the planet still cannot access or choose not to use. This may or may not be important to your results, but no matter how many people use online social networks, they are not representative of the entire population; more importantly, they may not represent YOUR entire audience. Twitter itself? Still (according to Henning) only about 7% of the online population, and even less of the overall population. Again, a fragment that may or may not represent the well-rounded sample you need.
The best quote (paraphrased) I pulled from Henning: “For best results you need to tie social media research into other results from outside social media.”
Online Personae May Not be True
Manila Austin of Communispace (the event’s host) pointed out that the opinions and emotions people express online may not be representative of their true selves. People censor themselves online, especially in public forums. I find that easy to believe, as anyone who belongs to private online forums with me know that I express myself much differently in those groups than I do in the public social streams (in which I try not to swear, and might pull critical punches a bit more than I do elsewhere). That “built-in focus group” I savored seven years ago? You may have to throw away those opinions as valuable feedback and capture these folks in a private setting, where they might be more honest and provide much more valuable feedback.
Don’t Let Your Expectations Limit Your Findings
Natasha Stevens and Michelle Bernardini of Visible Technologies (the program sponsor) brought up expectations. First, don’t go into research with preconceived notions of what the results will be because you will skew the results to fit your preconceptions. That sounds like basic market research practice but if social media is opening the doors to more companies conducting some sort of research, it bears repeating. Also, they brought up demographics; again, the social media world is not necessarily the whole world and the demographics that are present may not be the ones you need; their dominance does not necessarily mean others are not there. Take “Mom Bloggers.” They are a well-established, prolific and vocal group of people. They have built up such buzz that I found it necessary to capitalize “Mom-Bloggers” and put it in quotes (I’m not sure if that feeling influenced my hyphenation). Surely there are dads? Of course there are, but you still have to go find them if you are marketing to dads, rather than polling moms simply because you know you can find them.
It was refreshing to see the social media community get its claws into a meaty topic like research, as social media presents enormous opportunity to gain profitable insights, but also enormous opportunity to screw it up if you pursue it with the wrong frame of mind and expectations.
Photo credit: Derek Peplau on Flickr
For more from this event by other participants and attendees, most content will contain the “smb27” tag.