Digital Footprints in the Video Stream: Survey study of reflections on digital traces of media consumption and potential to use this for insights into well-being

Introduction & BackgroundNetflix now has a consumer base of over 230 million worldwide. During the pandemic, its customers watched 203.8 million hours of content daily, with their activity, content choices and preferences being continually logged. The digital footprint data amassed in this process underpins a symbiotic relationship between supplier and consumer. Black-box algorithms convert these logs into personalised functionality and recommendations, producing improved customer experiences while generating revenue for the business. Whether the consumer willingly accepts this trade-off or not, it’s now almost impossible to use online services without leaving digital traces. But how representative of an individual’s actual preferences and behaviours are these? What biases exist in such datasets? And to what degree are consumers cognisant of how these datasets are being used? Objectives & ApproachThis study surveyed participants to interrogate their understanding of the data Netflix makes available to its subscribers. The objectives were to explore their perceptions relating to the data collected about them and encourage them to think critically about their digital footprint. It was also the intention of the research group that participants feel a sense of empowerment / control over the data made available to them. UK-based participants were provided with instructions on how to access their viewing history (programme titles, dates of access) and invited to inspect it. 61 participants opted to donate their data to the study, along with responses to a survey reflecting their understanding of what they had retrieved. Relevance to Digital FootprintsWhile it may have been possible to work with Netflix to retrieve viewer data, by accessing via the participants instead, the researchers were enabling them to review and make informed choices about what they shared. One of the potential issues with this approach is that it provides an opportunity for participants to curate their data, should there be content that they would be uncomfortable sharing. Alternately, they may choose to withdraw from the study altogether based on what they see. While this has its drawbacks in terms of data inaccuracies and self-selection effect, it was felt important to the research team to prioritise the participants autonomy, encouraging them to be candid and share. If nothing else, it is hoped that by taking part in the study, there is the potential for participants to be inspired to t