Department of Management seminar Knowledge graphs as infrastructural devices: The case of Airbnb

With Dr Addie McGowan

Commercial knowledge graphs quietly arrange the markets made by the world's largest platforms, like Google, Meta, and Amazon. Yet despite being ubiquitous economization systems they have remained unseen by market studies scholars. This paper uses Airbnb as an empirical case to explore knowledge graphs as — quite literally — the dark side of platform markets, shedding light on their sociotechnical, epistemic, and market-making qualities. Drawing on market device (Muniesa et al., 2007) and market infrastructure scholarship (Kjellberg et al., 2019; Pinzur and Duterme, 2025), I ask how knowledge graphs make and sustain platform markets, and what this reveals about the existing relationship between these two concepts. Using a multi-method empirical approach spanning netnography, platform walkthroughs, and interviews with Airbnb engineers and hosts, I trace how Airbnb's knowledge graph qualifies, abstracts, and calculates cultural content as economic value, while also extending, standardizing, and politically negotiating its classifications at scale. These findings show the knowledge graph behaving as both device and infrastructure without fully meeting either concept's criteria. I therefore propose a new concept: an infrastructural market device. I hope this concept sheds light on the otherwise hidden data systems that platform markets design, maintain, and generate massive value with, advancing our analysis of platform markets.