Algorithmic Management of Work on Online Labor Platforms: When Matching Meets Control

Algorithmic control of platforms is used to control the behaviour of users on platforms to achieve the platform owners' corporate objectives.

“Algorithmic control of platforms is used to control the behaviour of users on platforms to achieve the platform owners’ corporate objectives”

Möhlmann et al. (2021) show bypassing and switching on labour market platforms as market-like behaviour and striking and embracing as organisational-like behaviour.

“Möhlmann et al. (2021) show bypassing and switching on labour market platforms as market-like behaviour and striking and embracing as organisational-like behaviour.”

Algorithmic distribution differs from algorithmic matching on online labour platforms and information diffusion among users of social media platforms by being a direct effect of algorithmic control.

“It also differs clearly from algorithmic matching on online labour platforms (Möhlmann et al. 2021) and information diffusion among users of social media platforms (Stieglitz and Dang-Xuan 2013) by being a direct effect of algorithmic control.”

Content creators adapt their behaviour by either avoiding or bowdlerising certain content, approaches that can even be interpreted as self-censorship.

“Our research shows that content creators adapt their behaviour (cf. Cram et al. 2020; Möhlmann et al. 2021; Wiener et al. 2021) by either avoiding or bowdlerising certain content, approaches that can even be interpreted as self-censorship.”

Findings on tensions have similarities with tensions reported in existing studies on online labour platforms, in that algorithm-versus-audience tension is similar to work compensation tension between uncertainty and repeatability.

“Our findings on tensions have similarities with tensions reported in existing studies on online labour platforms, in that algorithm-versus-audience tension is similar to the work compensation tension between uncertainty and repeatability, because content creators and Uber driver have difficulties understanding the logic of the algorithm (Möhlmann et al. 2021).”

Algorithmic distribution expands previous understanding of algorithmic control.

“In addition to the fine-grained distinctions within our aggregated dimensions, algorithmic distribution expands previous understanding of algorithmic control (Möhlmann et al. 2021; Wiener et al. 2021).”

Analytics-versus-decision-making-autonomy tension is similar to work execution tension between autonomy and supervision because preferences of analytics limit autonomy.

“Analytics-versus-decision-making-autonomy tension is similar to the work execution tension between autonomy and supervision (Möhlmann et al. 2021) because the preferences of analytics limit autonomy.”

The paper provides in-depth insights about monetary control and perceptions of CPM-related fluctuations that extend existing concepts of dynamic pricing.

“We also provide in-depth insights about monetary control and perceptions of CPM-related fluctuations that extend the existing concepts of dynamic pricing, which is dependent on time, place, and display (Asdemir et al. 2012; Choi et al. 2020; Möhlmann et al. 2021).”