Stealth Marketing

- Another Form of False Feedback

As Professor Strahilevitz points out, in using the reputation system, we should note that there are some people motivated to provide false feedbacks.  When he talks about such false feedbacks, he focuses on cases where valuators provide unduly low evaluations due to their discriminatory views.  Ryan’s example of “ratemyteacher.com” may be another example of such a falsely negative evaluation of the subject.  However, we should also beware of another type of information distortion, which may be as problematic – falsely positive information, or stealth marketing. 

- Intentionally Dishonest Reviewers

According to Wikipedia (or precisely its chache, due to the blackout), stealth marketing “is a subset of guerrilla marketing where consumers do not realize they are being marketed to.”  The word itself can include a wide range of activity, but let me take an example from a recent scandal involving the Japanese largest restaurant reviewing site called “Tabelog”.  It was reported that some reviews at Tabelog were actually posted by advertising agencies (Tabelog denies its involvement in the activities by these agencies).  Tabelog is an anonymity-based website, and we cannot tell which reviews are generated in the course of such marketing backed by restaurants.  Even if many of the reviewers are sincere as Professor Strahilevitz says, there may be those intentionally dishonest.

- Algorithm and Counter-algorithm

Shortly after the news, a geek created a website (http://crazyworks.jp/stlg/) which has a function to “guess” which restaurants on Tabelog are employing stealth marketing (please understand that restaurants with red texts in the column on the right are suspects of stealth marketing).  This website uses a very simple algorithm.  It basically regards a restaurant as a suspect, if (i) it receives more than 5 high-rated reviews, and (ii) many of those reviews are posted by reviewers with few past review posts (implying that the accounts might have been newly created for the purpose of marketing).  This website is a good example of an algorithm to exclude outlying reviews that are likely to be distorted.  However, we should note that counter-algorithm can be also invented, and such an invention can be encouraged especially when economic motivation is involved.  In this regard, such an algorithm should be held secret when false information is generated by commercial motivation, so that the counter-algorithm cannot be created.  On the other hand, the algorithm should be disclosed in case of a peer review at workplace, so that racist employees will learn to evaluate their minority colleagues equally in the hope of having their opinions more highly weighed.

- Fighting Stealth Marketing

Falsely positive information should be excluded as well as falsely negative information.  However, even if law prohibits such stealth marketing, it may be difficult or costly to enforce it.  Given that the reviewing website is anonymity-based, to detect a money backed reviews are far more difficult than to detect or prevent identity theft.  As Ryan suggests in the conclusion of his post, giving up the anonymity might lead to a solution (stealth marketing will be literally impossible).  But disclosing their names may discourage reviewers from posting reviews.  Alternatively, to scrutinize the reliability of each reputation system can be a stopgap measure.  Maybe we should not too much rely on the reviews on Tabelog, while we might be able to rely on the personal information disclosed in LinkedIn to some extent (given that the users are using the website with their names for the purpose of networking, to place accurate information there might be important to gain a job opportunity, or obtain professional trust from their “connections”).  Conversely, as Professor Strahilevitz suggests, in designing a reputation system, it is also possible to create a system encouraging participants to provide accurate information.

 

Comments

[Noah Yavitz]

I noticed this NYT article on the phenomenon of the "intentional dishonest reviewer" in the US: http://www.nytimes.com/2012/01/27/technology/for-2-a-star-a-retailer-gets-5-star-reviews.html?hp . The article also briefly mentions FTC rules requiring disclosure of a "connection between a merchant and someone promoting its product."

[Yohsuke Higashi]

Thanks. Personally I feel the www has become similar to the conventional media dominated by the commercialism. This may not be an arena of discussion among people on equal footing.