Today I'm happy to post splendid video and interesting article of Annie Pettite, PhD is the Chief Research Officer of Conversition Strategies, who, in my opinion is one of the most prominent Social media researchers.
Original artical is posted on scribemedia.
At Conversition, I’ve heard over and over again that companies are telling their employees do something, anything, in the social media space. With limited guidance, some people turn towards social media monitoring, others to customer relationship monitoring, and still others turn to social media research. Many struggle to differentiate between these offerings and simply go with whichever vendor is first in line.
As researchers ourselves, we understand the value that true market research brings to the table. True market research can come out of surveys, out of focus groups, and, no surprise here, out of social media data. Though social media monitoring can let you know what individual people are saying about a brand, social media research applies scientific principles to that data such that valid and reliable inferences can be made.
One of my favorite examples of correctly applying science to social media data relates to the use of Twitter, a social media tool that appeals to an arguably ‘different’ sort of person. As researchers, we strictly adhere to rules that say data must reflect the population to which you plan to generalize. As such, we know that even though twitter comprises only 7% of internet users, their voices cannot count for 60% of the opinions simply because that is how the data collection happened. Sampling, weighting, standards, norms, and other strict research processes are what place social media research into the market research toolbox.
Another of my favourite examples relates to the seemingly simple task of data quality. Whereas a survey can simply say “Coke” or “Apple” or “BP,” social media research needs to apply extremely strict and complicated data quality processes to ensure that the wrong data is not collected. Clients would certainly not wish to confound their datasets with opinions related to cocaine (coca-cola), apple pie (Apple computers), or blood pressure (BP).
What’s next on the horizon for social media research? Quite simply, predictive analytics based on social media data have arrived and are rapidly gaining momentum. The only reason market research is so important is that it helps users determine which products or services will thrive or fail. Knowing why something is failing right now is certainly useful and allows for real-time corrections and improvements to be made. But, in the long run, predictive research allows users to identify the hits and misses before massive quantities of money have been put into the products. Which television shows will be hits if a small tweak is made? Which shows should simply be pulled off the air? Which shows should be developed?
These are the sorts of things we already do on a regular basis with survey research. Decades of experience have helped us to discover how to create and combine variables to predict the winners of elections and consumer voted contests. We know how to combine past behaviour and intention variables to accurately predict the future. Now we are learning how to do those same things with social media research. We are creating the datasets that will be used to accurately predict the next flavour of chips, the next President, and the next American Idol.
