How Gamification Affects Crowdsourcing: The Case of Amazon Mechanical Turk

This paper, a 2019 article in Eludamos, argues that paid microwork platforms such as Amazon Mechanical Turk (AMT) represent a distinct form of IT outsourcing rather than a generic instance of crowdsourcing. Because the term has become overly broad, the analysis narrows its focus to paid microwork, in which labor is fragmented into small tasks, costs are reduced, and requesters retain full control over accepting or rejecting submissions. The primary purpose of these tasks is to produce, analyze, and process data, much of it used to train machine learning systems.

The central thesis is that gamification, through mechanisms such as points, leaderboards, and qualification tests, is used to motivate this workforce and increase productivity. An experiment cited in the paper found roughly 10 percent better accuracy and substantially lower per-image costs when gamification was applied, without any increase in worker pay. The paper situates these findings against labor data, reporting median hourly earnings of about $4.65 for US workers and $1.77 for Indian workers.

From these findings, the paper concludes that such gamification constitutes "exploitationware," a design approach that recasts productive labor as leisure and thereby enables the exploitation of an underpaid, on-demand workforce. The argument draws on literature addressing outsourcing, supply chains, and the evolution of IT labor models, though its conclusions rest primarily on AMT as a case study in the microwork economy.