Crowdsourcing business models : focusing on the crowd-labor industry and the implications for management and markets

Abstract

Purpose: This thesis analyzes the Crowd-Labor phenomenon, a subset of the Crowdsourcing industry where users of the related online platforms post tasks/projects and other users work on those tasks, usually in exchange for a payment. This work documents the current status of the platforms operating in this industry, providing new information regarding the numbers and trends. Its second objective is to understand how those companies are organized, what features they possess and how those features are related across different types of platforms. Methodology: Data collection regarding seventy-seven (77) characteristics from fifty-one (51) platforms. The characteristics are about the platforms themselves, their operations and the features they offer to their users. That was followed by an analysis of the data, and a grouping of certain related characteristics (for example, the sum of the number of available languages on the platform) and a correlation analysis to understand which types of platforms exist and what kind of platforms obtain that best performance. Findings: The analysis revealed that there are clusters of platforms based on the type of tasks/projects available on those platforms. Industry characteristics related with performance were analyzed, namely the existence of a forum, APIs, open challenges, the possibility of login & register using Facebook, fix payment fees for contractors, a leaderboard, the existence of multiple languages, internal exams for contractors to get certifications, tracking quality mechanisms and the possibility of project owners only paying when satisfied. Automated features (APIs and internal exams/certifications) stood out as a new positive performance differentiator for this recent industry, which is an original literature contribution originated from this thesis. Practical use: This work presents the current state of the Crowd-Labor industry, its benchmarks or industry standards, users’ motivations and a fact based opinion regarding its future, creating new knowledge that could be particularly useful for researchers, academics, crowdsourcing initiative owners, crowd-Labor users, entrepreneurs and investors. Limitations: An important limitation is that some of the answers to the characteristics used to analyze the Crowd-Labor Platforms were not made public by the platform owners, which didn’t make possible to capture the full picture of some of these platforms. However by studying 51 platforms, the collected data offers statistical evidence in the form of correlations that are statistically significant, which support the conclusions drawn from the analysis.

Date of Award	14 Feb 2014
Original language	English
Awarding Institution	Universidade Católica Portuguesa
Supervisor	Juan Andrei Villarroel Fernandéz (Supervisor)