Does My Opinion Count?

Does it matter I do not like Aishwarya Rai and love chocolates? Does it matter I like brand X, read Mills& Boons and that I am 33 years old ?

As a consumer, I wonder- am I important to a marketer?  Do I fall into a segment definition that would make my opinion on a brand/ category / topic matter for that marketer to do something about me? Is my opinion on facebook, part of the quantitative insights that make the brand or the aha’s that make up the insights?

These questions are interesting to me as a researcher.

Primary research is clear that there is quantitative or qualitative research. Quantitative research picks up a population frame from a universe, samples it and generates quantitative analysis of the brand/ category/ topic. Qualitative analysis takes a small sample and uses them to understand a topic in more detail and possibly generate a hypothesis from it.

Social media can do both.  Basis the opinions generated by millions of people of any topic of interest, we can generate the quantitative analysis for the brands. By listening to a few influential people talk about the brand we can help generate hypothesis for further testing.  That is good news, but how can that be done?

We at EmPower Research have learnt that we can borrow some principles from our well established primary research cousins.  Let us take an example of a research project that needs us to measure a brand or categories performance in social media.  Ideally, we would look at all conversations on a brand or category and then analyse what is being said. Inspite of all the tools in existence in the market;  to mine all relevant social media conversations is an exercise fraught with errors. So we do the next best thing, we set a population frame using search strings and then pick out a sample for our reading and analysis. This is the STRANDS methodology that we have used across multiple clients successfully.

One of our clients wanted to understand social media conversations around severe allergic asthma. We defined an exhaustive search string and used this across multiple social media channels to mine all relevant content- population frame set.

We discovered that stakeholder conversations fell into a patient lifecycle of pre diagnosis, diagnosis, treatment and lifestyle. We used the stratified sampling methodology to ensure adequate representation across social media channels; strata representation from each of the patient lifecycle stages;  adequate representation of stakeholder and lastly the brands representation.

We then used the sample to read and provide quantitative analysis and insights for the client. This might sound simple, but believe me, we have taken years to get it right and convince our clients about the validity of this research approach. Is this fool proof, of course not? Remember social media research pertains to opinion of those people who participate online only, for starters.

What are the areas that social media turns out to be a more powerful than traditional research.  To understand consumer segments (what do moms think, how do teens perceive charity), topics that consumers do not react favorably to or forced to react favorably when questioned(sustainability, fertility anyone?). That is a topic for another blog…

- By Priya Venkataraman