Antitrust Implications of Food Service Aggregator Owned Cloud Kitchens

0
3036

-Aditya Kumar Singh and Hardik Malik*

As delineated in the recent case of In Re Prachi Agarwal & Ors. v Swiggy (“Swiggy case”), Food Service Aggregators (‘FSAs’) like Swiggy and Zomato operate in the “App based food delivery platform” market with India being the relevant geographic market.[1] But lately, these FSAs have branched out to become market participants by registering self-owned ‘cloud kitchens’ on their own platforms. A cloud kitchen or ‘ghost kitchen’ is a restaurant that cuts down on all the overhead costs, eliminates the typical brick and mortar structure and only offers delivery or takeout services. This article highlights the anti-competitive ramifications that follow the neutral FSAs’ (in this case, Swiggy) decision to become a market participant.

Dominance in the Food Delivery Market

The first step towards analysing a case of abuse of dominance is to establish that the concerned enterprise is dominant in the relevant market. The Competition Act, 2002 (“the Act”) by way of S.19(4) lays down the various factors to be considered for an inquiry into the dominant position of any enterprise. The first among these is S.19(4)(a) i.e., market share of the enterprise. At the end of 2018 an independent data intelligence platform placed Swiggy’s market share at 48%,[2] but Swiggy categorically denied being a dominant enterprise and opposed this ‘static analysis’ by relying on Zomato’s acquisition of Uber Eats and the dynamic nature of this market. In 2018, CCI accepted Swiggy’s contentions.[3] But, despite Zomato’s acquisition of Uber Eats, and the dynamic nature of the market, Swiggy’s market share kept rising, cornering 60% of the market at the beginning of 2020. Thus, as per the market share, Swiggy is clearly the dominant enterprise. 

The second factor to be considered is S.19(4)(h) i.e., barriers to the entry of new enterprises. The food delivery market entails high capital cost of entry as the FSAs need to maintain a huge fleet of delivery executives in every city to ensure on-time delivery. If a new participant wants to gain a substantial market share, they would need to compete with these two FSAs, and currently, Zomato boasts a fleet of 150,000 drivers while Swiggy has 130,000 drivers. Further,  a majority of the restaurants registering on the FSAs are small scale businesses that depend on the FSAs for all their orders, with the FSA charging them 15-20% commission per order. Thus, the restaurants would only want to register on a platform with a large consumer base, while the consumers would only want to use an FSA that offers a wide variety of restaurant options. This network effect prevents a rapid growth of any new entrant in a market where the existing players already have close to 200,000 restaurant partners listed with them

One might still raise the question – can’t players enter the market in one city or one state and compete with these FSAs locally? This is answered using the example of Amazon’s recent entry in the Food Delivery business, where it started food delivery in a select few areas of Bangalore in May, 2020. However, even as of March 2021, Amazon delivers food to just 62 out of the 270 pin codes in Bangalore. This goes to show that even an e-commerce giant like Amazon with its money, brand value, and data analytics capacity could not cover even 25% of the one city they started functioning in over a year ago. Further, as per a report from Motilal Oswal, to reach even this figure, (of close to 25%) Amazon had to reduce its commission per order to half of what Swiggy and Zomato charge. Thus, it is clear that this market has a severe logistical and financial barrier to entry. As a result any entrant, if there even is one, would require years to set up shop and become profitable. 

The third and final factor is section 19(4)(j) i.e. market structure and size. After Zomato’s acquisition of Uber Eats and Food Panda shutting shop in India, the Indian FSA market consists of just Zomato and Swiggy, making it a virtual duopoly. In the recent Telecom sector market study, the CCI while elaborating upon the rule of three has accepted the idea that any mature market needs three main competitors to ensure healthy competition. Further, it also established that the exit of even one leads to a virtual duopoly, which is detrimental to competition.[4] This is because the decision of either of the two remaining parties will then have a sizeable impact on the market. Thus, in light of the aforementioned factors it is safe to say that despite Swiggy’s contention to the contrary, it does hold a dominant position in the app based food delivery market.

Abuse of Dominance

It is argued that Swiggy is using this dominant position in the app based food delivery market to enter into another relevant market i.e. the food service market, a conduct that falls within section 4(2)(e) of the Act. As per the CCI’s decision in MCX Ltd. v NSE, the requirement for establishing a case for abuse of dominance under section 4(2)(e) is to identify two distinct relevant markets and establish a connection between them.[5] As shown above, the two relevant markets in this case are the food service market and app based food delivery market. Both these industries are directly connected as the former serves as an input for the latter. Further, the manner in which the dominance in one industry is being used to enter and gain unfair advantage in the other can be categorized into Quantitative Bias and Qualitative Bias.

[1] Quantitative Bias

In the Swiggy Case, the informant had alleged that Swiggy is misleading the customers by charging higher prices than the prices charged by the restaurants themselves.[6] In response, Swiggy revealed its ‘Merchant Terms of Use’ agreement which mandated that restaurants registering on its platform maintain price parity across all points of sale, whether online or offline.[7] In doing so Swiggy might have averted the immediate charge but it opened itself to an entirely separate line of enquiry. Theoretically, the concept of price parity seems user friendly but a contextual analysis of the same suggests otherwise. The ramifications of such an arrangement are that brick and mortar restaurants that include overhead costs like cutlery, wait staff, rent etc. in their prices for physical diners, are forced to charge the same amount from consumers placing orders online. Whereas the cloud kitchens which only have an online presence, don’t have to account for any of the aforementioned overhead costs and can price their products at a much lower amount. An illustration of the ramifications of this ‘Merchant Terms of Use’ is how it enables the FSA owned cloud kitchens to offer heavy discounts.[8] While restaurants are forced to charge the same amount online and offline, the cloud kitchens can easily mark up their price and make it similar to the offline restaurants. After this, they can offer deep discounts to bring the price down to the actual level, making the deal seem much more attractive to the consumer. This enables the FSA to offer discounts of up to 50% and still make profit on the transaction.

One might still raise the question – how feasible is it to charge a lower amount online than offline? Especially after paying the FSA a commission of 15-20%. Admittedly, not all restaurants would want to charge a lower amount but the restaurants should still be given the choice to decide their own prices. Further, online sales are a great way to publicize the restaurant. Therefore, these discounts should not be seen in a vacuum, as some part of the money allocated for advertisement of the restaurant can be used to offer discounts. This is because discounts incentivize people to try out new restaurants, enabling such restaurants to expand their customer base.

Therefore, this ‘Merchant Terms of Use’ qualifies for abuse of dominance under section 4(2)(a)(i) i.e. imposing unfair or discriminatory condition in the sale of goods. As the condition of across platform price parity practically dictates the price that the restaurants can charge for their own products, a price that directly benefits the sales of the FSA owned cloud kitchens to the detriment of such restaurants. This conclusion also finds support from the findings of the European Commission in the case of Amazon asking e-book publishers to sign an across platform price parity agreement. The Commission felt that such an agreement would grant Amazon an unfair advantage and would be detrimental to the entry and expansion of competitors.[9] This led to Amazon discontinuing the agreement.

 [2] Qualitative Bias

FSAs like Swiggy offering private labels are financially incentivized to provide preferential treatment to their cloud kitchens, to the detriment of their competitors. In the food industry the FSAs act as an intermediary between the consumers and the restaurants giving them the power to undermine competition by pursuing largely undetected qualitative biases like: (a) manipulation of search results and prominent placement of their own products; (b) asymmetrical access to information relating to consumer behavior and patterns from the platform.

[2.1] Platform Neutrality

The allure of a specific or generic online marketplace lies in its ability to collate all possible sellers on a user-friendly platform. The consumers can then evaluate and compare the quality and prices of various products and their retailers before deciding. In reality, consumers are not perfectly rational. They are limited by their imperfect knowledge of the products offered and usually do not spend time analysing every product and maximizing satisfaction. In such a scenario, the FSA’s classification of products (eg: the “certified safe” tag during the pandemic), placement of products, and algorithmic responses to searches by consumer become extremely important in increasing user traffic, likability, and subsequent purchase of products.  This is especially true for the food delivery market which offers close substitute goods and the qualitative factors like the platform placement and search response play an instrumental role in the sales.[10] Thus, a product’s visibility is directly proportional to a consumer’s propensity to purchase it.

But the biased intermediary is uncompetitive from an economic (as demonstrated above) as well as legal standpoint. On the legal front the actions of these FSAs prop up two major concerns i) under section 4(2)(a)(i), the imposition of discriminatory condition in sale of goods; and ii) under section 4(2)(b)(ii), technical development to the prejudice of consumers. 

[2.1.1] Discriminatory Condition in Sale of Goods

The CCI has unambiguously stated that product placement and search algorithms are entirely capable of creating disparate user traffic and subsequent purchase, which in turn disincentives other sellers from continuing to sell on these platforms and/or misleads the consumer about possible alternatives. Further, even the European Commission in the case of Google Search (Shopping) was of the view that prominent placement of your own product works to the detriment of your competitor’s sales. This led to the commission fining Google 2.42 billion euros for preferential treatment leading to non-pricing abuse of dominance.[11]  Therefore, it stands to reason that by offering their own cloud kitchen prominent placement[12] and better classification the FSAs are discriminating between their customers (partner restaurants) in terms of conditions for sale. This conduct squarely falls under section 4(2)(a)(i) i.e. imposing discriminatory condition in sale of goods.

[2.1.2] Technical Development to the Prejudice of Consumers

Dynamic and subjective search algorithms are the boon and bane of our antitrust regime. They are the manifestation of how some data-driven and AI-based technologies perpetrate anti-competitive actions. It is a boon because an antitrust regime is not intended to hamper innovation or promotion of commercial interests by the intermediary. The intermediary chooses its search algorithm and product placement with the singular aim of improving consumer experience and generating consumer traffic. This will obviously involve subjective decisions of what furthers the intermediary’s commercial interests. But the problem starts when the intermediary becomes a market participant. In its recent e-commerce study the CCI analyzed the algorithms for various platforms, describing the algorithms used by the FSAs to select its ‘Top Picks’ as a ‘black box’. This was primarily due to the complete lack of transparency about the parameters considered for deciding the ‘Top Picks’. Resultantly, because of not being aware about the parameters, restaurants are rendered incapable of improving upon the same. Thus, this algorithmic bias can be said to have satisfied the criterion of limiting or restricting technical development to the prejudice of consumers under section 4(2)(b)(ii) of the Act. The FSAs being a two-way platform also treat the restaurants as their customers. This unilateral opaque application of an algorithm designed by the FSAs,  that judges the restaurants on the basis of unknown parameters works to their prejudice by depriving them of objective choices. As the restaurants are oblivious to the parameters, they cannot make an informed choice as to whether to continue with the practices that resulted in the low rating or change the same to obtain a higher one.

Furthermore, even if one of these consumers wanted to file a complaint against the FSA for being anti-competitive, this analysis of platform neutrality would be predicated on the presence of extensive technical know-how about the algorithms. This is because the informant is supposed to establish a prima facie case of bias before any investigations can be ordered.[13]But the algorithms being a ‘black box’ and the informant not having the statutory powers or expertise of the commission, makes the discovery of impropriety onerous. This is precisely what has prevented an in-depth investigation into Google’s search bias on Play Store. The CCI was of the opinion that mere screenshots depicting biased search responses were insufficient evidence to establish a prima facie case. Thus, the opaqueness of algorithmic biases prevents identification of self-preferencing or other such anti-competitive behavior.

[2.2] Asymmetrical Access to Information

The CCI has also observed a general complaint amongst restaurants that the platforms never shared critical customer data like individual feedback, general preference for cuisine or dishes[14] etc. It was the commission’s finding that the FSA can use its role as an intermediary to mine information like demand, customer preference etc., and then as a market participant, come up with “cloud kitchens in high demand food categories”.[15] A manifestation of this can be observed from the menu of ‘The Bowl Company’, a Swiggy owned cloud kitchen, which offers a mix of Indian, Chinese, Thai, Mughlai, and Italian dishes with the restaurant not fitting in any of the usual cuisine categories. Thus, it is argued that an abuse of dominant position in the food delivery market to establish a foothold in the food service market is apparent. 

Conclusion

The spirit of the Indian antitrust regime lies in protecting competition and not the competitors. Admittedly, the emergence of cloud kitchens has revolutionized the food service industry to the disadvantage of traditional brick and mortar restaurants, while still being within the contours of healthy competition. But, the anti-competitive concerns arise when the Food Service Aggregators that regulate the “neutral platform” simultaneously enter the market as a participant. Currently, the Indian antitrust regime does not seem equipped to adequately deal with the algorithmic biases creating disparate user traffic and subsequently affecting sales. As seen in the case of Google the burden of proof, even for a prima facie case, seems too high to be satisfied by small-scale individual restaurants with no technical know-how or access to the algorithms. Thus, the CCI needs to take it upon itself to investigate this abuse of dominance and level the playing field.

* Aditya Kumar Singh and Hardik Malik are third year students at the National Law School of India University, Bangalore.


[1] In Re Prachi Agarwal & Ors. v SwiggyCase No. 39 of 2019, ¶ 3.

[2] Ibid, ¶4.

[3] Ibid, ¶15-16

[4] Ibid, ¶60.

[5] MCX Stock Exchange v National Stock Exchange, Case No. 13 of 2009, ¶10.78.

[6] In Re Prachi Agarwal & Ors. v SwiggyCase No. 39 of 2019, ¶ 8.

[7] Ibid, ¶ 14.

[8] CCI-Market Study on E-Commerce Sector in India, 2020, ¶107.

[9] Richard Whish & David Bailey, Competition Law (OUP, 9th ed., 2018) 731.

[10] CCI-Market Study on E-Commerce Sector in India, 2020, ¶107.

[11] Richard Whish & David Bailey, Competition Law (OUP, 9th ed., 2018) 730.

[12] CCI-Market Study on E-Commerce Sector in India, 2020, ¶ 58-60.

[13] Surendra Prasad v. CCI & Ors. in Appeal No. 43 of 2014,  ¶22.

[14] CCI-Market Study on E-Commerce Sector in India, 2020, ¶59, 71.

[15] Ibid, ¶ 59.

LEAVE A REPLY

Please enter your comment!
Please enter your name here