Random Sampling

Social media can be a ‘voice of the customer’ goldmine — monitoring comments and discussions, looking for consumer mentions of a company’s brand or product.

A big motivator for this work has been crisis management.  It’s no wonder.  At this point, no one can deny the power of social media to quickly spread and amplify experiences with a company, product, or brand – good or bad.  Two examples include the song on YouTube about United Airlines losing a customer’s guitar and the Motrin ads that set off a firestorm among ‘mommy-bloggers’ who felt that Johnson & Johnson was suggesting that their babies are a ‘pain.’

But is social media monitoring “real” research?

I think the answer depends on how you define “real research.”

Many researchers have avoided this practice and some have been pretty outspoken in their criticism. Usually, they question the representativeness of the information.  Because the information isn’t collected from a probability sample*, it’s not useful for anything more than collecting some stories.

(* Neither is most “reliable” quantitative research, but that’s a story for another post.)

Many skeptical researchers use the Motrin/Moms ad frenzy as a prime example of why social media monitoring can’t be trusted.  As a result of online outcry against the ad, Johnson & Johnson swiftly pulled it.   However, critics say, follow up quantitative research concluded that the advertising was not offensive to the vast majority of moms.  This, according to them, is irrefutable proof that using social media for research is unreliable and leads to bad business decisions.

Conversely, if you define “real research” as any data collection and analysis activity that provides intelligence about the market to support business decisions, then, yes, social media monitoring is very much “real” research.  Like every other research technique, it all comes down to finding the appropriate application for the tool.

Will social media monitoring deliver a statistically valid sample of the market? No, not as it’s currently done. But that doesn’t invalidate the approach.

As I see it, social media monitoring helps marketers do two very valuable things:

• Identify influencers who are talking about the company, brand, or product.
• Understand the zeitgeist of the market’s opinion leaders.

Because much of social media monitoring focuses on blogs, social networks, and Twitter, the comments being found are mostly one-to-many communications.  So, social media monitoring is very good at finding the people with big soapboxes who are talking about the company.  Once found, companies have the opportunity to engage these people, which can go a long way to ensuring that the information being put out is accurate, or at least that the company’s perspective might be considered.

The tone and tenor of the social media chatter may or may not be representative of the broader market, but it is going to be fairly representative of the influencers.  For the most part, what the influencers are talking about today is going to become accepted truth among the broader market sooner or later.  To me, not paying attention to that discussion is just irresponsible.

I tend to have a broad view of what research should be.  To me, any activity that is primarily concerned with understanding what is going on in the market is something that should interest researchers. That absolutely includes social media monitoring.

It seems to me that the research industry’s narrow focus on the survey research model (and social science-based qualitative approaches) has closed researchers’ minds to emerging market data sources. This is to the industry’s detriment.

How about you?  Do you perform any sort of social media monitoring?  Do you think researchers should be involved in this work?  Let me know!

Next week, I’ll talk more about applications of social media to research.  Don’t miss it, subscribe to the RSS feed now.