Jeremiah Owyang from Forrester writes on his blog about study on "trustworthiness" of various sources of information (unfortunately, traditional platforms like newspapers or magazines were not measured in the study). According to the study, 84% of North American consumers trust "opinion of a friend or acquaintance who has used the product or service", compared to only 30% who trust "an online review by a blogger". Now, while the study itself is interesting (attitudinal research and Word of Mouth measurement are my venues), I would be careful with interpretation of its results. This is the comment I left on Jeremiah's blog:
"One should carefully interpret results of this study, because it does not take in consideration such factors as prior distribution of exposure to different media/sources of information. In other words, what you think is indication of “source trustworthiness” may be in fact simply reflectin of levels of “reach” of that source in general population. We communicate with our peers, friends and family everyday, and in various settings. “Word of mouth” has a reach of nearly 100%. TV has a penetration of nearly 95%. These sources are available and accessed by nearly everyone, so more people refer to them as to “most trustworthy” sources. Online blogs, on the other hand, are read by only 8% of US population (and only 11% of US Internet users). Chat rooms and discussion boards are used by merely 15% of Internet users. As a result, these sources were mentioned by less people.
Now, I am not trying to defend blogs or social media. I am just poiting out that it is necessary to take in consideration prior distributions of these sources. Marketers often limit their research to simply reporting distributions, whereas one should also look at how variables of interest related to other factors. Reach and frequency is soooo analytics 1.0! Traditional media have been using these metrics for a long time, and now it looks like digital media now social media are repeating their mistake. Think in terms of niche-marketing: the most attractive segment is not the one that is bigger, but the one that is more likely to use your services. So I would like to see first how trust in each source is related to likelihood of purchasing product or services after seeing a review."
With regard to trust vs. reach, I completely agree with Jeremiah and other researchers mentioned in his post.
What I would do differently? I would look at the conditional distributions: ratio of number of people who trust opinions and product reviews by their friends or acquaintances and actually buy the products to number of people who regularly discusses products and services with their friends or acquaintances. Similarly, look at the proportion of people who trust product reviews by bloggers out those who regularly read blogs, etc. etc. I would also look at gender/age/socio-economic differences in "trustworthiness" scores and compute correlations with various behaviors and attitudes.
