Competitive Behavior in the Attention Economy - The Case of Twitch Streamers

Wollborn, Philip; Reindl, Matthias; Ehrmann, Thomas

Zusammenfassung

Recent research argues that entrepreneurs who generate their main income through online digital platforms are subject to specific power asymmetries. The platform firms can often single-handedly impose changes that can have a strong impact on the success of entrepreneurs using these platforms, which lead to the term of “platform-dependent entrepreneurs” (PDEs). These dependencies can be especially crucial within the “attention economy”, meaning community-focused business like social media platforms. On these platforms, PDEs (e.g. influencers) often are “price-takers” with an explicit price of zero as their core content is typically free access by other platform users. It is only through the attention generated by these PDEs that they can build a brand and monetize their content, with typical examples being sponsorship/advertisement deals, merchandising, exclusive premium content, or paywhat- you-want options. One particularly interesting platform in this matter is the online live-streaming platform Twitch. On Twitch, users (streamers) are able to broadcast a diverse portfolio of live content, typically featuring themselves, while interacting directly with their audience. As live-streams are typically free to watch for viewers, streamers have only two strategic options at hand to maximize monetization: what to stream and when to stream. These two options are related to the actual structure of competition: both the content to be streamed (what) and the time of streaming (when) can be viewed as an optimal (re-)action to the (re-)actions of all the other streamers active on the platform. In this context, we provide empirical results regarding the competitive behavior of streamers as well as an example for the risks and chances of PDEs face through decisions by the platforms they operate on.We show that streamers tend to pursue temporary clustering in time and content, similarly to the Hotellingmodel, though not all components of these strategies seem to actually increase audience size and subsequently income generation.