Changes to scoring algorithms used by commercial operators of online platforms may violate antimonopoly law

Introduction

On June 16, 2022, the Tokyo District Court ruled that “Kakaku.com”, the operator of an online restaurant review and search platform, had abused its superior bargaining position and violated the law. antimonopoly by inappropriately altering its restaurant rating algorithm.

This article provides an overview of the decision and a discussion of the key factors that led the Court to determine when changes to an algorithm would violate antimonopoly law. The case provides a useful guideline for trading platform operators who are considering altering their algorithms.

Please note that the content of this article is based on media coverage, commentary and other public sources since the text of the decision has not yet been made public.

II. Case summary

Kakaku.com operates the online restaurant rating and search platform, “Tabelog”, where customers can search for restaurants using a variety of criteria, including location, cuisine type and price range. Additionally, customers can register and leave reviews and ratings. Restaurants can also register their business on Tabelog for free. For an additional fee, restaurants have access to additional tools to raise their restaurant’s profile and reach a wider audience. Restaurants that pay certain additional fees are also placed preferentially in users’ search results. Tabelog is one of the largest and best-known online restaurant review and search platforms in Japan, with around 830,000 restaurants listed and 93 million monthly users as of June 20221.

Hanryu-mura, a chain of barbecue restaurants, filed a lawsuit claiming that Kakaku.com changed the algorithm used to calculate restaurant ratings in May 2019, which caused Hanryu’s restaurant ratings to drop. -mura.

Hanryu-mura claimed that Kakaku.com secretly tweaked the algorithm to lower restaurant chain ratings to trick the company into paying higher fees to have their restaurants appear higher in ratings. search results. Hanryu-mura sought an injunction against the use of the modified algorithm and JPY 639 million in damages.

Although the circumstances are still unclear, it was reported that Kakaku.com leaked its algorithm to Hanryu-mura as part of the legal process. This is a rare occurrence because algorithms are generally considered highly confidential information and their disclosure is usually fiercely contested.

The Tokyo District Court denied the plaintiff’s request for an injunction against the use of the modified algorithm, but awarded the plaintiff 38.4 million yen in damages after ruling that the modification of the algorithm by Kakaku.com was an abuse of his superior bargaining position.

III. Abuse of a superior bargaining position

(1) Superior Bargaining Position

Under antimonopoly law in Japan, it is prohibited for a business operator (Party A) to abuse a superior bargaining position over another party (Party B) by setting or changing the terms of of a transaction or by carrying out a transaction which interferes with fair competition. . Unlike the abuse of dominance prohibition in the EU, Party A need not have a dominant position in the market; rather, it suffices that Party A simply has a relatively superior bargaining position to Party B.

In this case, the Court found that Kakaku.com had a superior bargaining position, stating that Hanryu-mura had no choice but to accept even disadvantageous demands from Kakaku.com because he would otherwise face a major problem in the management of its restaurants if the restaurants could not continue to be paid members of Tabelog. This discovery is perhaps not so surprising given the important role that Tabelog plays in the restaurant industry in Japan.

(2) Completion of a transaction

A key issue in this case was whether the change in the scoring algorithm constituted disadvantageous “implementation of a transaction” by Kakaku.com. Although restaurants can become paid members of Tabelog, the ratings of each restaurant are not part of the terms and conditions of this paid membership. Tabelog displays the ratings of not only paying members, but also free members and even non-member restaurants that have no business relationship with Kakaku.com, and the ratings do not go up or down depending on whether the restaurant is a member paid or not. Kakaku.com argued that rating a restaurant did not constitute a “transaction”.

On this issue, the Japan Fair Trade Commission (the “JFTC”) submitted an amicus brief in response to the Court’s request. The amicus brief states that, although Tabelog rating is not part of the contract with paid members or free members, it is a service provided as part of a transaction and is at least the “implementation of a transaction”. The JFTC gave the following reasons:

(i) Tabelog score is an indicator that shows how many ratings have been collected from users at that time for the restaurant;

(ii) Restaurants can become a paying member of Tabelog to update their information and increase their visibility to consumers through “Profile Registration” and “Access Enhancement” to attract more customers; and

(iii) The rating of the restaurant is increased by increasing the number of user ratings and comments, thereby attracting more customers.

However, since the text of the judgment has not been made public, it is not known on what grounds the Court concluded that the modification of the algorithm constituted the disadvantageous implementation of a transaction.

(3) Obstruction of fair competition

The other issue in the case was whether the algorithm change was an impediment to fair competition. The JFTC’s amicus brief stated that the following factors should be considered in determining whether the establishment and operation of the algorithm in question interferes with fair competition:

(i) the overall content of the algorithm applied to restaurant ratings and the circumstances of the changes (eg, what factors are taken into account, how often the factors are reviewed and changed);

(ii) how, when and for what range of restaurants is the algorithm defined and operated (including whether or not there is prior consultation with restaurants);

(iii) whether the algorithm is likely to remove restaurant autonomy; and

(iv) the extent to which the algorithm is harmful to restaurants.

In making its decision, the Court may have concluded that the algorithm change in question interfered with fair competition based on the JFTC criteria above.

IV. Conclusion

This decision is particularly noteworthy in that it shows that an unfair algorithm change could violate antimonopoly law, and disclosure of the algorithm may be required as part of a legal proceeding. Platform operators should carefully consider all factors set forth by the JFTC when changing their rating algorithm, including the reasonableness of the change, the degree of impact on the counterparty, and the extent to which the operator will conduct prior consultations. Further information and insight into this case will be available once the text of the decision is made public. Additionally, both parties appealed the case to the Tokyo High Court. Particular attention should be paid to future developments.

Sharon D. Cole