Data mining techniques prove to be an inexact science, as results from content measurement firms vary significantly and turn out questionable. San Francisco, California-based Telephia recently had to explain results from its survey on mobile TV in the United Kingdom. Mobile-video experts in the U.K. backlashed at the results, which stated that one-third of U.K. mobile TV users watch BBC1. The conclusion appears questionable as only one handset in the country, Virgin Mobiles Lobster, can accommodate the show. Telephia stood by its results, explaining that the survey asked users what they have watched covering off-deck sites, streaming and video clips. Another issue emerged when Telephia and its competitor M:Metrics came up with contrasting results about the mobile-gaming industry. M:Metrics reports that activity was flat for several months in both the United States and European markets, whereas Telephia reports a 53 percent increase in downloads from January to March 2006. The two companies use different methodologies in their studies of mass-market handsets. Telephia is inclined to monitor the monthly bills of consumers, while M:Metrics uses consumer surveys. Both methodologies have pitfalls that could lead to inaccuracies. Bill monitoring can become especially inaccurate in cases such as family plans. Consumer surveys largely depend on how good the respondents recall their activities.