Social Media: The End of Conventional Market Research?
A New Market Analysis Methodology
I just finished reading about a San Francisco startup called Topsy Labs. This company searches posts on social media websites and uses this information to detect forthcoming trends. For example, it picked up a lot of tweets from people who said they were cancelling their Netflix subscriptions and used that information to predict a drop in Netflix’ stock price. There are other new companies doing similar work, for example, WiseWindow and Derwent Capital Markets, a London-based boutique investment company running a hedge fund.
The work of these companies is based on sentiment analysis in which the chatter on Facebook, Twitter, blogs and other social media is analyzed and used to predict stock movement, market trends, product acceptance, competition and other factors that is the purview of classical market research. It is a bit early to predict the demise of classical market research, but it will certainly be impacted in a significant way.
Will the Statistics Texts be Rewritten?
I spent more than 30 years working for market-research based companies. We were called “industry analysts”. Setting aside things like focus groups, the principal data-gathering MO of such companies is interviewing. We did face-to-face interviews, phone interviews and online surveys interviews. We tried to do enough of them to be “statistically significant.” (It turns out that the statistics books state that 50 interviews of a homogeneous population will produce results with ±10% accuracy with 90% confidence.). Nonetheless, obtaining 50 great interviews wasn’t easy and was very expensive.
I don’t know what the sentiment analysis companies are charging or going to charge for their services, but, since the process is highly automated, it could be relatively inexpensive. In fact, it could ultimately put the market research firms out of business or force them to change their their existing business models. In reply, representatives from those firms are touting the limitations of sentiment analysis, but the quotes I’ve seen aren’t very convincing.
Conclusion: Sentiment Analysis and Spin-offs Will Have a Huge Impact
In conclusion, I predict that data gathering by scanning social media and other web-based information will have a huge impact on the way market research will be done in the future, and will provide a precision unmatched by conventional techniques.