Market Research + Sentiment Analysis = New Insight
Q & A with Bill Tuohig, J.D. Power & Associates
Posted October 24, 2012
Bill Tuohig is Senior Director of Social Media and Text Analytics Products at J.D. Power & Associates. He will be speaking on "Using Sentiment Analysis to Make Net Promoter More Actionable" at the up-coming Sentiment Analysis Symposium, October 30, 2012 in San Francisco.
Bill is one of four authorities who graciously agreed to respond to a series of questions exploring the role of sentiment analysis, text analytics, and emerging social-intelligence technologies in support of next-generation market research. Read his responses to questions from Seth Grimes and then return to the full set of interviews..
Q1> What impact is sentiment analysis having in the market-research world, whether applied for surveys or to social media?
Bill Tuohig> For quite some time, Market Researchers have avoided or ignored unstructured data captured through surveys. Text analysis with sentiment assessment is causing more and more researchers to mine the data sets they've already collected to unlock untapped value, as well as include more open end questions than before. Also, one of the big challenges in survey research is declining response rates. In an attempt to address this, some researchers are re-thinking how they design surveys, using open end questions to create a more streamlined survey experience, instead of an endless list of closed end questions.
We find researchers are starting to use social media analysis as a new part of their research efforts as well. In some cases it is used to "prime the pump" for focus groups, in others even replace focus groups. As methodologies advance for working with this type of data and this type of scale, I expect to see social media analysis play a greater role in the research portfolio.
2> Could you describe one or two things you or your clients have learned, via automated text/sentiment analysis, that you wouldn't have discovered otherwise?
Bill> In general, clients often learn the "why" behind the "what". It's one thing to know someone's not satisfied with, say, the nav system in their vehicle... but what is it about that system they didn't like? What was frustrating? What do they wish was different? It's these types of insights that we commonly find to help clients inform their action planning, and in some cases, uncover issues they didn't know about, or solutions they hadn't considered.
3> How consistent are research findings from social media and from surveys?
Bill> It really depends on what you are studying. When it comes to measuring things like advocacy or likelihood to recommend, social media MR often tracks quite closely with surveys. However, when you are measuring things like customer satisfaction or detailed quality scores, we sometimes see significant differences. From our perspective, the key is to understand what each can be used for and when to use them. If you want to track consumer reactions to something like a product launch in terms advocacy and awareness, social media analysis can be a very valuable tool, providing real-time intelligence that surveys cannot provide. However, if you want to find customers who have had a specific experience with your brand or product in order to measure customer satisfaction, surveys allow you to control the sample in ways social does not. So the key is understanding when to use each - neither one is "the best" at everything.
4> Are there quality concerns in social-media MR, for instance due to not having representative sources or due to language and usage irregularities? And are there special advantages in social-media MR?
Bill> Yes, again it comes back to really understanding the data you are working with. There is a lot of noise out there in social media, and it is not easy to filter it out to get to the true data you want to work with. It can be difficult to isolate specific segment of people with confidence based on automated algorithms today - they are getting better, but just aren't there yet. And, without solid research methodologies for setting up queries, running quality checks, etc. you can get very different results depending on which analyst does the work. So there are a number of challenges to working with this data, both in terms of the processes, standards, and data characteristics.
That said, social media analysis does provide some unique benefits that can be invaluable to an organization. For example, one is speed: Organizations commonly tell us they need to be able to make faster business decisions, and in many cases they are willing to trade of exact precision for speed. Social media MR can provide solid directional insights much faster than other traditional approaches, and sometimes that's the difference between winning and losing. Another is scale: when you tap into a population that large, you can quickly find segments that would otherwise be very hard and expensive to find through traditional techniques.
There are certainly many other strengths to Social Media MR that make it an important tool to have in the market research portfolio, and I would encourage Researchers to get started now in learning about this data, how and when to use it, and to get some practical experience under your belt so you start to understand how you can apply it in your business.
Meet Bill Tuohig and other market research innovators at the Sentiment Analysis Symposium, October 30, 2012 in San Francisco. For now, return to the full set of interviews.