Experience: 

Digital platforms exert a monopoly by using opaque search algorithms

Many dominant digital-based companies exert a monopoly on their market (e.g. Facebook, Google, Amazon, Apple iTunes, GooglePlay and App Store, eBay, etc.). This monopoly bears on the distribution platform of the tangible or intangible goods being sold. Access to the end customer of this category of goods is only possible through this platform. It is a "natural" monopoly, due to the technical and cost advantages of size in a network, which favour the largest player and reinforce their dominance on the market. It is also a worldwide monopoly, due to the fact that the Internet eliminates the obstacle posed by physical distance.

Digital platforms control access to the final customer and exert pressure on their suppliers by using opaque algorithms to determine which products or documents will be displayed to the customer first. Such algorithms can be subject to discretionary changes of parameters, so as to arbitrarily and selectively favour or impede one supplier or the other.

Ideas: 

All websites that sell products or select information should include a “fair” search algorithm.

Defining such an algorithm is a difficult task. However, a “concentric search”, whereby all relevant results are first displayed in a random order (this randomness being controlled by third party inspection of the source code), with an easy option to refine search within these results, could be a first, viable, technical option which merits further investigation[1].

When such a transparent and fair search option is available, the suppliers suffer no economic pressure from the digital platform, because they cannot be discriminated against in a discretionary manner. They can generate more value added, provide better, more innovative products, and better, fairer working conditions for their own workers and for those of their suppliers. A “fair” search algorithm also brings advantages in terms of freedom of information and cultural diversity, by ensuring that the range of information or digital content being displayed to the final user is neither biased nor limited in favour of the advertisers’ interests, or their political or cultural preferences.

The website would be free to keep its existing search engine, and would thus propose two alternative options to the user:

• its current model, where an opaque, potentially biased, algorithm provides the information to the end user "for free"[2], whilst favouring the interests of the platform and those of its submissive suppliers
• the transparent, fair model mandated by the regulation, where this "fair" search algorithm is implemented and certified by an independent third party. Here, the user can be requested to explicitly pay[3] to use the service – but with the added certainty that s/he is obtaining the results (products or documents) that s/he (and not the digital platform) has chosen.

[1] While it is beyond the technical competence of industriAll European Trade Union to provide a full solution to this problem, we hope to show, through this proposal, that at least one simple and feasible solution exists – albeit unsophisticated. The existence of such a solution proves that: (1) early implementation is feasible, and (2) further research makes sense. This discussion may be an indication of how central the “search within results” feature is to retain control on one’s search strategy.

[2]The costs being borne by the real paying customer, i.e. the supplier paying the advertisement and all the other fees, discounts, etc...) “If it’s free, you’re not the customer: you’re the good being sold”.

[3]Either with on-line micro-payment solutions (that are yet to be developed, or are emerging, such as the EU-based SatisPay, and have been long-deserved to counter-balance the hegemony of the advertisement-funded model), or via a subscription model with a "remember me" function.