3.2.2 Free and subsidized data use Research on free and subsidized data use was conducted in over two weeks in Yangon, Myanmar in July 2016. The team conducted focus group discussions with a total of 63 mobile phone users as well as informal interviews with corporate stakeholders and street-side vendors. Males and females from SEC A to E from urban and rural Yangon were selected. All 63 respondents were Internet users, i.e., users of mobile data. At the time of research, Myanmar had a range of zero-rated data offerings—the most prominent being Facebook Free Basics (offered via MPT, the incumbent operator), and Free Facebook and Viber text messages (offered by new entrant Telenor). All operators offered a number of other subsidized or promotional data packages. The research found that respondents did not use or know of Free Basics content other than Facebook. While some 40 respondents had used Free Basics, only four knew of free content available on the platform beyond Facebook and Messenger. Respondents described different behavior on the two main zero-rated promotions in this study. Many respondents stopped using Free Basics because of user-experience frustrations: the absence of photos and video on the free version of Facebook, slow data speeds and the process of switching back and forth between paid and free content. Those who continued to use Free Basics used it—alongside paid data—for limited purposes. Many respondents used Free Basics primarily when their top-up balance had run out, as a means to keeping in contact with others before topping up. Respondents who used Telenor Free Facebook and Viber increased data consumption on the promotional packages available. Several rural respondents began watching video content on Facebook for the first time because of the promotion. Most used their entire 150MB free quota each day, and many paid for additional data. Thus, the promotion served as an ‘on-ramp’ to paid data consumption, although not to the ‘open Internet’: users paid to continue using Facebook. Telenor Free users were content to remain within the ‘walled garden’. Respondents also used zero-rated promotions as a strategy to manage data-costs. These respondents used one of several strategies on Free Basics: using free Messenger exclusively, choosing to read posts on free Facebook, or limiting certain Facebook activity to free mode, in particular checking Facebook groups. Some users of Telenor Free stopped once they reach the daily cap, while others continued using Facebook by switching to MPT Free Basics. User perception mattered: if users understood that they were using free content, i.e., they see the ‘garden walls’, then they understand they are not using the ‘open Internet’. The limitations of Facebook on MPT Free Basics served to highlight these garden walls: respondents knew when they are using zero-rated content and when they were not. The same cannot be said for Telenor Free. Several users—from both urban and rural focus groups— described the free 150MB Facebook allotment on Telenor as general-use data. Perception and visibility aside, users did not remain within the ‘walled garden’. Most active zero-rated-content users also used other Internet services, including Google, news websites, and apps. Respondents describe following links from within Facebook to external websites. Exiting the walled garden was more common among urban respondents, but most rural respondents who used zero-rated content also use other applications online, commonly BeeTalk or Clash of Clans. Given the choice between zero-rating, limited data to access unrestricted content, or unlimited access to open content but at slow speeds—three widely proposed models for spreading access—respondents preferred unrestricted content. Respondents expressed frustration at slow Wi-Fi and data speeds. Frustrations with