Automatic Pricing Mechanisms in Competition Law: The Turkish Competition Board’s Buybox Decisions

Pricing Algorithms in Competition Law 

Today, digital platforms—on which consumers rely to meet many of their daily needs, ranging from online shopping to transportation and accommodation to energy consumption—make extensive use of advanced algorithms in the background. With the digitalization of commercial life, it is increasingly observed that platforms resort to algorithms and artificial intelligence tools to improve their pricing models, monitor consumer behavior, and track market trends. On one hand, algorithms serve various functions, such as improving production processes, encouraging innovation, reducing costs, and facilitating market entry. On the other hand, they enable consumers to make more informed decisions by allowing comparisons based on price, quality, and preferences, thus reducing search costs. In this context, pricing algorithms that are capable of processing comprehensive market data and balancing supply and demand can support more efficient market functioning and help consumers meet their needs at lower costs. While algorithms offer significant efficiency gains for both undertakings and consumers by enabling the delivery of better and more personalized products and services, they also carry the risk of being used in ways that reduce competition and harm consumers. Specifically, the way algorithms are used can allow businesses to come to unspoken agreements without needing a written contract or face-to-face meetings, which raises worries about unfair competition.. Among these concerns, pricing algorithms—commonly used in e-commerce to help undertakings to track the pricing strategies of competitors—stand out prominently. These algorithms increase market transparency and facilitate mutual monitoring among competitors, potentially enabling them to reach consensus on elements that influence strategic decision-making processes, particularly price. Examples of anticompetitive collaborations facilitated by algorithms include the determination of resale prices through automated pricing systems, hub-and-spoke cartels that involve information sharing among undertakings using the same third-party software, and inter-firm cooperation that emerges through self-learning algorithms in the absence of explicit coordination. As a result of such practices, algorithms can give rise to tacit agreements among undertakings, ultimately leading to price rigidity, diminished competition, and slowed innovation—all of which can negatively impact consumer welfare.

Automatic Pricing Mechanisms Implemented by Trendyol and Hepsiburada under the Board’s Buybox Decision 

In its Buybox decision, the Turkish Competition Board examined whether the automatic pricing mechanisms used by e-commerce platforms Trendyol and Hepsiburada violate Law No. 4054 on the Protection of Competition. The automatic pricing mechanism is a system that automates the price transitions of sellers who offer their products for sale on e-commerce platforms within the framework of certain rules. This system aims to reduce the time and transaction burden required for sellers to monitor and update prices on the platform. However, the mechanism is essentially designed to automate the competition between sellers for a place on Buybox. In the Buybox, different sellers selling the same product are gathered under a single heading; when a product is searched by users, the product of the seller who wins the Buybox is displayed first and ranked on the product page according to their Buybox score. Being featured in the Buybox is therefore crucial for sellers, as it directly influences their visibility to consumers and significantly impacts sales performance.

Sellers can update their prices through the automatic pricing mechanism offered by marketplaces, and they can determine how much below or above the price according to their own strategies. In this context, the three basic rules that sellers can determine are ‘Stay Below Buybox Price’, ‘Match Buybox Price’ and ‘Stay Above Buybox Price’. In the ‘Match Buybox Price’ rule, unlike the other two rules, it is not an option for the seller to enter any information in terms of amount or percentage; the seller is only expected to enter the ‘Lowest Applicable Price’ information.

In light of the above, the Buybox algorithm is taken as a reference in the automatic pricing mechanism applied by marketplaces. Considering that being featured in the Buybox is important for sellers in terms of visibility and sales, the parameters affecting the Buybox algorithm applied by marketplaces were also examined within the scope of the Board decision. The marketplaces argued that, the algorithm does not solely focus on price; benefiting from the automatic pricing rule does not necessarily guarantee that a seller will win the Buybox, as the algorithm operates based on multiple metrics. It was also emphasized that the use of the automatic pricing mechanism is not compulsory for sellers, as sellers may set prices through alternative means, and that the system primarily serves to save time and effort, particularly for those managing a large inventory. Consequently, the Board concluded that the automatic pricing mechanism is not mandatory for marketplace sellers.

As a result of its assessment of the case file, the Board acknowledged that automatic pricing mechanisms are not inherently restrictive by object in all circumstances and may, in fact, generate consumer benefits. In this regard, given that the use of the automatic pricing mechanism is not mandatory for sellers, prices are not solely determined based on the Buybox, the Buybox itself is calculated by the platform using variable parameters, there is no meeting of minds among sellers, and the system is designed to function in a differentiated manner, the Board concluded that the mechanism does not constitute a manifest and serious infringement in terms of its operation. On the other hand, the said mechanisms, offered to sellers by different marketplaces and functioning very similarly, are considered to pose a risk of causing price rigidity in the market, especially if the ‘Match Buybox Price’ algorithm is widely used. Although these systems are offered within the framework of individual and vertical agreements, when many sellers start to use the same algorithm, they may cumulatively cause anticompetitive effects at the horizontal level. In other words, sellers may indirectly harmonize each other's prices through algorithms without a conscious agreement. In this context, it is decided that the commitment proposals submitted by Hepsiburada and Trendyol to address the competitive concerns raised in this context are suitable to address the anti-competitive concerns.

Commitments Submitted by the Parties

To address the competition concerns raised within the scope of the file, commitments have been offered by the undertakings. Within this framework, the rule ‘Match to Buybox Price’, which poses a particular risk of leading to coordination, will be removed; and it has also been committed that under the existing rules ‘Stay Below Buybox Price’ and ‘Stay Above Buybox Price’, sellers will not be allowed to define rules that could result in the same outcome as the ‘Match Buybox Price’ rule, either in terms of percentage (0%) or amount (0 TL). Additionally, it has been committed that specific information regarding the automatic pricing mechanism – such as whether the mechanism is being used or how frequently a particular rule is applied – will not be provided to sellers; the use of this mechanism will not be mandatory; and no incentive equivalent to an obligation will be offered. Furthermore, it has been stated that the automatic pricing mechanism will not provide any advantage in the buybox algorithms, and that sellers using this mechanism will not receive higher scores in the buybox ranking solely for this reason. 

Conclusion

Pricing algorithms offer features that can positively impact competition by enhancing the competitiveness of undertakings and enabling consumers to compare prices more effectively. However, the way these algorithms are used may give rise to anti-competitive concerns, as they can facilitate tacit collusion among undertakings without the need for any written agreement or physical interaction, and maintain such collusion. In this context, as demonstrated in the Competition Board’s Buybox decision, it is crucial to carefully examine the effects of algorithms on competition and to take necessary precautions to preserve a healthy competitive environment on digital platforms.