How platform sellers work around the algorithm
18 April 2021
Core Insights: Future of Work
By Gerardo Patriotta
Algorithms have become part of our daily lives with digital platforms such as Amazon, eBay, Airbnb, Tripadvisor and Uber using them to gather online reviews from a huge number of customers who are geographically dispersed and operating independently.
Reviews are compiled into a score which represents ‘the wisdom of the crowd’ and can be harnessed to provide confidence in the platform.
In a recent study we interviewed 77 large-volume, high-performing business sellers who were conducting their business on eBay and generating positive ratings above 95 per cent.
The initial motivation for our research was an interest in how online market platforms were affecting the online sellers’ work practices. Our interviews focused on the general experience of online selling, interactions with buyers, the relationship with eBay’s staff, as well as the benefits of receiving more detailed sales data.
What we discovered was surprising. We had been expecting our subjects to report positive experiences of online selling. Instead, we found a pervasive sense of frustration caused by negative reviews and impersonal algorithmic procedures. Online evaluations generated a sense of vulnerability among sellers who described their experiences as a “sword of Damocles hanging over our heads”, “prison-like” and “Kafkaesque”.
How do we explain this puzzle? Our findings pointed to asymmetric power dynamics. Whereas in traditional marketplaces like the high street, buyers and sellers enjoy equal status, online selling distorts this balance. Thanks to eBay’s algorithm, online buyers are aggregated as a crowd whose individual actions produce collective evaluations, while sellers are left as isolated individuals.
The online platform represents a three-way power structure where power is distributed unevenly between the seller, the buyer, and the platform operator.
Buyers have the power to evaluate anonymously without being evaluated. These evaluations have a direct impact on sellers for two reasons. First, they are public. A seller with very high ratings will attract more buyers than a seller with low ratings. Second, eBay uses these ratings to manage sellers and could even, in extreme circumstances, ban the seller from the platform.
Caught in the crossfire between buyers and an algorithmic ‘boss’, sellers must respond by improving their service, ensuring speedy delivery and guaranteeing goods meet a required standard.
Online customer evaluations prompt sellers to develop a practical knowledge of the algorithm. They regain control by developing communication, customisation and relational strategies based on their superior knowledge of the platform. In doing so, they create their own rules, forcing buyers to abide by them, and blocking them if necessary. In other words, they develop practices to work around the algorithm instead of working for an algorithm.
Why are algorithms so important?
An algorithm is a procedure or formula for solving a problem, based on conducting a sequence of programmed actions. It is based on sophisticated, albeit simplified, modelling of situations in which individual behaviours can be measured and aggregated into a score. It is on the basis of this simplified representation of reality that algorithms assess employee behaviour in terms of appraisal, motivation and reward.
Organisations have been characterised as machines. They are technical instruments designed to achieve predefined goals with efficiency, precision and control. Algorithms support calculative decision making and increase the speed of decisions: they process large amounts of information very quickly and suggest optimal decisions. In doing so, they improve organisations’ ability to monitor their environments.
But algorithms have a dark side that is linked to power and surveillance. Online customer evaluations have created a new form of employee monitoring through a coalition between buyers and the platform owner, who join together in evaluating the seller. This coalition is mediated by algorithms that operate anonymously and are programmed by platform designers to act on behalf of the platform owner. In other words, algorithms are carriers of power.
Do algorithms represent a step change for the future of work? To be sure, algorithms help organisations amplify their rationality by leveraging the ‘wisdom of the crowd’. They prompt sellers to improve their service based on customer ratings. But algorithms are programmed to act on the basis of precise instructions and automated procedures. They do so quietly and anonymously. One cannot negotiate with an algorithm: its evaluations offer no right of reply.
The power of the algorithm lies in its ability to generate asymmetrical relationships. The intrusion of power asymmetries in online settings, generally associated with independence and autonomy, may provoke strong negative feelings and attempts to escape or game the algorithm.
But our research has shown that once online sellers understand these ground rules, they can learn to use algorithms to their advantage and work around them.
Curchod, C., Patriotta, G., Cohen, L. and Neysen, N. (2020) "Working for an algorithm : power asymmetries and agency in online work settings", Administrative Science Quarterly, 65, 3, 644-676
For more articles on the Future of Work sign up to Core Insights here.