In a recent article by F. Annie Pettit in the MRA's Alert! Magazine (http://www.mra-net.org/alert/), the author discusses Twitter and potential market research applications. The article (titled "Can Marketing Research Really be Conducted Using Twitter?") is worth reading for several reasons.
First, social media and the digital explosion is an undeniable trend, and it can be harnessed for qualitative (not quantitative) research purposes. Twitter could clearly be used as a part of a qualitative feedback loop for corporations or institutions with a desire to constantly track customer perception and experience. In the marketing and research industries these systems are often referred to as EFM or "enterprise feedback management". Twitter/Tweet monitoring and tracking could make a useful complement to the hosted online communities that many consumer brand companies are building. Communispace (http://www.communispace.com/) and VoVici (http://www.vovici.com/) are two hosted online community providers that leap immediately to mind. If they aren't already, I would expect these two firms to develop social media conversation monitoring as a tandem product.
Second, the author clearly notes that (a) it is impossible to make Twitter data nationally projectable and (b) it is unlikely that it could ever become a quantitative tool. She clearly notes that, "a major drawback reflects the lack of a quality sampling method, a core requirement for obtaining valid and reliable research results." Inevitably, there will be companies over the next few years that will claim to have a conversation analytics offering built upon a social media application that is a statistically projectable quantitative tool. This will annoy real researchers immeasurably.
Third, Pettit notes that on larger subjects or widely used products, an enormous amount of qualitative data could be mined. With advances in text analytics software, research and marketing teams could identify fast moving trends by looking for words or phrases that pop overnight. For example, a travel destination like Dubai could track tweets that utilize standard travel terms and Dubai. A daily or weekly sampling of this datastream could then focus on aspects of the consumer experience and look for patterns, problems and opportunities. Large CPG companies and service-based companies (airlines, amusement parks, etc.) could find this analysis useful.
Finally, monitoring and analyzing conversation streams like Twitter is a passive, non-interventional type of qualitative research. Recently referred to as "listening outposts" in the market research community (see Surinder Siama's April, 2009 article in Research World), there is something to be said for the purity and authenticity of this type of unprompted qualitative data.
This is just the beginning of a much larger trend in market research. Watch for it.