Haaga-Helia’s AlgoAmmatti: Algorithmic Management and Professional Growth in Platform Economy project has been at the forefront of understanding and guiding how AI should be used in a responsible way on digital labour platforms. This article summarises the key findings of the project.

Researching algorithmic management

The AlgoAmmatti project explored the concept of algorithmic management, specifically AI-based work facilitation and management practices in platform economy companies and the effects of these practices on the day-to-day experience and professional growth of platform workers. The aim was to develop a worker experience centred model of algorithmic management and professional growth.

In practice, the study consisted of three work packages and comprehensive empirical data, mainly semi-structured interviews (n=19), qualitative questionnaire (n=258), netnographic forum post analysis (n=830), document analysis (n=14), six LEGO Serious Play workshops (participants n=36), six focus group interviews (participants n=18), and a questionnaire (n=103).

Based on our research, many platforms in Finland use design features that can be generally characterised as algorithmic control. For example, individual platform workers’ – or service providers’ – visibility on platforms is of paramount importance. As visibility is often controlled by the platform’s AI algorithms, those working on the platform may, for example, deliberately underprice their services or agree to additional work in the hope of good reviews – or in fear of bad reviews. For some platforms, it may also be a matter of ostensibly maintaining the activity of a user account for fear of being removed from the platform.

Two key suggestions for improving digital labour platforms

A key finding of the AlgoAmmatti project is that those working through labour platforms have little transparency or influence on how the AI-based functionalities of the platform work in practice. Based on the empirical data collected in the AlgoAmmatti project, we therefore call for more transparency. We see two main ways to do this.

First, we recommend platforms to introduce user assemblies composed of platform users and other key stakeholders to support decision-making (Tuomi & Ascenção 2023). Based on the principles of participatory design, a user assembly would provide a structured and transparent way for the voice of platform users to be heard, increasing trust and commitment to the platform. As new features are introduced to the platform – for example, updating the platform’s fee-setting algorithm (Wolt 2023) – it would be important for platform users to have a genuine say in the direction the platform develops. Critically, for a user assembly to work, it should be genuinely representative of users, meaningful, and have genuine power, for example veto power over certain types of decisions.

Second, we see the need to develop the way work is done on platforms towards a life-cycle approach. In principle, platforms do not offer career progression opportunities in the same way as a regular paid job does, be it through greater responsibilities or salary progression. Platforms also have a certain interest in being passive in offering career progression opportunities, as many platforms have a high annual churn rate of workers and a queue of new willing candidates. For example the logistics platform Wolt has publicly stated that it has around 5,000 people working weekly as couriers in Finland, but that “there are around twenty thousand people in the queue to become couriers” (Pankakoski 2023).

From the perspective of transparent algorithmic management, it is crucial to consider, for example, how to define an inactive courier account, i.e. the logic for shelving old accounts and opening new ones so that there is enough work for everyone. Overall, a shift in thinking that emphasises the lifecycle of platform work, rather than seeing those working on the platform as economic units or as inspiration for product development and marketing communications (Wolt 2023b).

For example, platforms could support the professional growth of platform workers through different types of training. Platforms could also facilitate the transition periods associated with platform work, such as moving from one platform to another, from working on platforms to paid employment and from paid employment back to platforms, and from working on the platform to working for the platform.

Project ends, research continues

The platform economy is a new form of work, which, according to research and growth forecasts, seems to have arrived in Europe and Finland to stay (European Council 2022; Viitanen & Eskola 2022). In part it enables, in part it restricts work. AlgoAmmatti project examined algorithmic management practices of platform companies and the effects of these practices on the perceived everyday lives of those working through platforms.

The project resulted in the development of a framework for transparent platform economy development and algorithmic management as part of the platform worker lifecycle. This crystallises the direction and ways in which platform economy companies operating in Finland should be developed based on the information gathered in the project. We see Finland as a unique testing ground for experimenting with new ways of working and an opportunity for thought leadership based on democratic values in relation to international markets.

Although the AlgoAmmatti project will end in 2023, research and development work on the platform economy will continue. We hope that the project publications we have compiled will provide interesting reading at the intersection of user-driven platform economy development and algorithmic management.

Editing: Marianne Wegmüller

Picture: Shutterstock