Latausnopeuden ja muiden muuttujien vaikutus käyttäjäkokemukseen katsottaessa sisältöä verkkopohjaisilla videoalustoilla
Raunio, Teo · Aaltodoc (Aalto University) · 2020
This master's thesis studies and evaluates the impact of loading speed and other variables on the user experience of viewing content on web-based video platforms. The number of other variables researched was not limited in the beginning of the research, and the available data determined the selection of beneficial variables. The selected method for approaching the problem is secondary data analysis and the data is provided by two separate analytics services and which were collected by a company developing video platforms. The research focuses on a single video platform and the goal was to achieve results that could be generalized to other video platforms, and used in the development of new video platforms. The main motivation for the research was the fact that there is previous work on the subject of loading speed, but not notably for the purposes of a video platform by taking into account the additional loading needed to initialize the video playback. As the results of the analysis were transformed as attributes that affect the UX, four metrics were defined to measure their impact on the user experience. The metrics were usability, speed, attractiveness and reliability. In addition to loading speed, the other selected variables for the analysis were buffering, interruptions, errors, happiness score, device type, browser type, connection type, streaming type and live DVR playback, which resulted in many coherent results. As these results were evaluated according to the UX metrics, the loading speed, along with other variables, had a clear impact on the user experience in the given context. Of the four metrics, speed and reliability had the largest number direct factors affecting the UX, whereas usability and attractiveness required more interpretation of the results. For future research, it is recommended to analyze and compare the results of two platforms or conduct an experiment that is believed to improve the UX, while the data is being collected, to see its impact. Also, tools for inspecting loading speed in more detail are recommended.