A blog post on Twittown today recommended  Five Wickedly Clever Ways to Use Twitter. These are useful tips but I want to add a note of caution to Rob’s fifth recommendation to use Twitter for focus groups.

For one thing, it’s unlikely “millions” of people will be talking about your product unless you’re an airline and one of your planes has just emergency-landed in the Hudson River. Unless you’re a household brand, the Twittertalk about your brand will be relatively muted. Sample size is critical to evaluating the reliability of results. Smaller samples (e.g., fewer users, or mentions) will be less reliable.

Second, a successful focus group depends on having a fair representation of the demographic who interact with the product or brand. On Twitter, your demographic is restricted to, well, Twitter users. And they may not be a typical slice of your demo.

Third, moderating a real-life focus group takes a lot of training and skill. You need to know how to manage group dynamics (which may or may not be important to your findings), and how to guide the group towards revealing the kinds of information you want. Marketers have barely begun figuring how to do such things on Twitter.

So by all means use Twitter for focus groups, but bear the above caveats in mind or else you’re going to go wrong. Use results from Twitter in conjunction with other methods (surveys, feedback analysis) and the Twitter data will provide an additional perspective.