{"id":605,"date":"2010-07-14T01:00:29","date_gmt":"2010-07-14T05:00:29","guid":{"rendered":"http:\/\/www.yupingliu.com\/wordpress\/?p=605"},"modified":"2010-07-13T23:18:26","modified_gmt":"2010-07-14T03:18:26","slug":"quitting-behavior-is-social-too","status":"publish","type":"post","link":"https:\/\/www.yupingliu.com\/wordpress\/2010\/07\/14\/quitting-behavior-is-social-too\/","title":{"rendered":"Quitting Behavior is Social Too"},"content":{"rendered":"

When we talk about social networking or contagion effect, we are usually referring to getting good words out about us so that we can engender good will and gain additional customers. In other words, we often focus on positive behavior in the context of social networks. But just as positive behavior can be fostered through social networks, quitting behavior can be social too. When consumers decide to leave a company, even when no particularly negative word-of-mouth is present, that decision can still have a social rippling effect. Consider my recent quitting of Facebook<\/a>, for instance. That decision was bolstered by seeing 36000+ consumers who also signed up to quit on QuitFacebookDay.com<\/a>.<\/p>\n

 <\/p>\n

\"Exit\"
\n
Image by Loui Loui<\/a> from Flickr | CC 2.0<\/p>\n

To help understand how this social quitting behavior works, I would like to discuss the findings from a rare academic study on this subject. Published in the Marketing Science Institute’s<\/a> 2010 Working Paper Series, Irit Nitzan and Barak Libai from Tel Aviv University studied the effect of defection by friends on our own decisions to quit a company. If you are interested in the full report, you can purchase it directly from MSI<\/a>.<\/p>\n

Context<\/strong><\/p>\n

The study is based on the behavior of 853,643 customers of a major cellphone service provider in a Mediterranean country. The communication records and defection behavior of these consumers over the course of one year were examined. Social networks were constructed from the consumers’ call and text messaging records.<\/p>\n

Main Findings<\/strong><\/p>\n