[by Vashmath Potluri and Shubhranshu]
The authors are students of NALSAR Hyderabad.
Introduction
India’s e-commerce market has rapidly consolidated, with Amazon and Flipkart controlling over 67 percent of the market. While this dominance is often attributed to scale and logistics, the Director General’s (“DG”) 2024 investigation report reveals deeper structural concerns. Both platforms exercise infrastructural control over warehousing, logistics and algorithmic discoverability in ways that consistently privilege select sellers and marginalise unaffiliated rivals. Practices such as exclusive launches, fulfilment-linked visibility boosts and restricted consumer access reflect a broader pattern of selective gatekeeping. However, the Competition Commission of India (“CCI”) continues to assess such conduct under Section 3(4)(c) of the Competition Act, 2002 (“the Act”), treating the platforms as vertically aligned intermediaries. With August 2025 marking one year of the DG report and the CCI’s final order still pending, similar concerns have emerged against quick commerce platforms like Zepto and Blinkit. Therefore, the Amazon–Flipkart case marks a turning point for Indian competition law as it could shape regulatory responses to infrastructural exclusion across platform markets.
This article proceeds in two parts. Part I argues for an ex post reclassification of the conduct under Section 3(3)(c), treating selective infrastructural access as a form of horizontal market allocation. This would invoke a per se presumption of appreciable adverse effect on competition (“AAEC”) once coordination is shown. Part II offers a forward-looking, ex ante regulatory framework by drawing from global models such as the EU’s Digital Markets Act (“DMA”) and the UK’s Strategic Market Status regime (“SMS”), it proposes structural and behavioural tools to prevent infrastructural foreclosure at the design stage. Together, these approaches aim to restore open competition in India’s platform market economy.
Ex-Post: Establishing Horizontal Market Allocation
The DG’s 2024 investigation revealed that the exclusion on Amazon and Flipkart was not incidental; it was embedded in platform design. Both companies consistently privileged a small cohort of sellers, six on Amazon and thirty-three on Flipkart to be precise by providing them early inventory access, algorithmic prioritisation and subsidised warehousing. These advantages were especially visible during exclusive launches, where unaffiliated sellers were systematically denied access to high-demand stock keeping units, despite having similar operational capabilities. The DG’s conclusion that “no seller other than its preferred seller can survive” highlights that this was not sporadic favouritism but a deliberate exclusionary structure.
This exclusionary design is best understood as a form of “Hub-and-Spoke coordination,” with platforms acting as hubs and preferred sellers as spokes. While the sellers may not directly communicate, the platform facilitates alignment through observable and repeatable incentives. In CCI v. Coordination Committee of Artists, the Supreme Court held that tacit arrangements may constitute agreements when reflected in sustained, parallel conduct facilitated by a central structure. Here, algorithmic favouritism and real-time visibility generate constructive knowledge. Sellers observe which behaviours are rewarded, such as integrating with platform logistics or participating in exclusivity, and calibrate their conduct accordingly. This form of indirect alignment, stabilised by the platform and repeated across cycles, supports an inference of agreement under Section 2(b) of the Act.
At the centre of this arrangement is infrastructural segmentation. In digital marketplaces, search visibility, fulfilment logistics and promotional tools are not neutral; they are levers that shape competition. While Indian jurisprudence has not yet defined these as standalone markets, international regulators increasingly treat them as critical gatekeeping mechanisms. The European Commission, in its enforcement under the DMA, flagged Apple’s restrictions on interface functionalities, such as preventing developers from linking users to external purchase options, as materially distorting market access even in the absence of total foreclosure. Similarly, the OECD has recognised that platform-controlled tools such as ranking systems, algorithmic design, and fulfilment infrastructure can act as structural barriers by controlling visibility and consumer access. When allocated selectively, particularly during launch cycles, these tools replicate the exclusionary impact of classic market-sharing.
Recognising this infrastructural segmentation as a form of horizontal coordination is therefore a doctrinal interpretation grounded in functional realities. In FHRAI v. MakeMyTrip, the CCI held that algorithmic prioritisation of OYO, combined with suppression of competitors, amounted to an exclusionary agreement. The same logic applies here because algorithmic and logistical design choices made by Amazon and Flipkart repeatedly favour the same seller cohort. This enables parallel outcomes among competing sellers, driven not by direct collusion but by their mutual orientation to platform-curated incentives. Such structurally induced alignment occurs “in any other similar way” as contemplated under Section 3(3)(c), fulfilling the evidentiary threshold for coordination even in the absence of a traditional horizontal agreement.
Importantly, this interpretation also aligns with the second proviso to Section 3(3), which extends liability to entities that act “in furtherance of” anti-competitive agreements,” even if they are not engaged in “identical or similar trade.” By designing and enforcing exclusionary infrastructures, Amazon and Flipkart move beyond the role of passive intermediaries and become active participants in market segmentation. Classifying their conduct as horizontal market allocation ensures that Indian law can respond effectively to structural exclusion embedded in platform design.
Shifting the standard from the rule of reason to Per Se
Once Amazon and Flipkart’s conduct is reclassified under Section 3(3)(c) of the Act, the standard of liability undergoes a fundamental shift. Currently assessed under Section 3(4), such conduct requires a “rule-of-reason analysis,” where the CCI must affirmatively demonstrate that an agreement causes or is likely to cause an AAEC. This approach places the evidentiary burden on the regulator. In contrast, Section 3(3) adopts a “per se rule” because once a horizontal agreement with an exclusionary object is established, AAEC is presumed, and the burden shifts to the parties to provide compelling evidence of overriding pro-competitive justifications.
This presumption is more than a procedural shortcut; it reflects a structural understanding of platform markets. In these markets, tools like fulfilment access, algorithmic ranking and promotional placement are not ancillary. They define the terms of competition itself because when such tools are selectively allocated to a few preferred sellers, the result is not mere inequality but systemic distortion of competition. The per se rule recognises that exclusion embedded in platform design requires no further analysis of consumer benefit at the liability stage. Claims of faster delivery or smoother user interfaces may be relevant in vertical contexts but collapse in horizontal scenarios where foreclosure is built into the architecture of access.
In this context, Section 19(3) no longer serves as a threshold test of harm but as a narrow efficiency defence. Here, Amazon and Flipkart must show that the exclusionary conduct brings indispensable, pro-competition benefits that fully offset the presumed harm. The threshold is set high because efficiency alone cannot justify exclusion, especially when the exclusionary effects clearly outweigh any pro-competitive gains. This particularly fits in digital markets where infrastructural exclusion is long-lasting, path-dependent, and rarely corrected by market forces alone. Short-term consumer convenience cannot justify long-term denial of access to a strategic lever of competition.
A good example of this principle is the ruling of the European Commission in Google Shopping. Google regularly promoted its shopping service and demoted competitors, citing that it improved the user experience. The Commission rejected this because consumer convenience could not serve as a justification for the systematic denial of competitive visibility. The underlying issue was not whether it was beneficial to the user now, but the structural damage to access and contestability by rivals. This is specifically relevant to Amazon and Flipkart as by coordinating with a narrow set of sellers, they want to promote them by providing algorithmic advantages and prime search listing including prioritized fulfillment not through their merit, but through the architecture of the platform. Such omissions compound over time, entrenching advantages for insiders while leaving others without reasonable opportunities to enter the market
Viewed in this light, Section 3(3)(c)’s per se standard is not rigid but tailored to platform realities, where exclusion emerges not through explicit refusals but through design choices that shape access, discoverability, and fulfilment. While horizontal coordination often suggests express collusion, platform-driven alignment can achieve the same result through constructive knowledge. Sellers adjust behaviour in areas such as pricing, inventory, or exclusivity based on clear and repeatable rewards embedded in the platform’s infrastructure. This results in parallel outcomes that replicate market allocation. Therefore, Section 3(3)(c) would capture this form of structural exclusion by shifting the burden onto platforms to justify preferential infrastructural access. This interpretation ensures that competitive advantage flows from merit rather than discretionary platform design.
Ex-Ante: Regulation of Platform Market Algorithms and Infrastructure
The architecture of exclusion in the Indian digital markets has taken a new form that involves a reduction of direct contractual barriers to tacit and design-induced distortions. Platforms increasingly do so by concealing the black-boxed algorithms and biased execution and discriminatory onboarding processes that determine future access and visibility before causing immediate price and output effects. These exclusionary designs are present before being measurable and therefore, ex post enforcement is not sufficient. These developments have led to the adoption of ex ante regulation by the EU and UK as a means of regulating concerns of fairness and transparency in platform design. India should also do the same by amending the Act to deal with design based exclusion. This part suggests the three-pronged road map of regulation.
-
Mandating Algorithmic Transparency
Algorithmic design is not the passive aspect of platform design any longer; it is a strategic lever that determines visibility and fulfilment priority, as well as market access. The CCI must respond not merely through reactive scrutiny after complaints arise but by embedding transparency as a structural regulatory mandate. To that effect, the CCI must make use of its rule-making authority in Section 64(1) and 64(2)(h) to provide binding Algorithmic Disclosure Regulations on systemically important digital platforms.
Based on the Article 5(1) of the EU Platform-to-Business (P2B) Regulation, these regulations are supposed to require periodic disclosures of
(i) the fundamental parameters used for ranking of the sellers and products, (ii) their relative weighting, particularly with respect to affiliation and fulfilment integration, and (iii) whether private labels or preferred sellers systematically benefit from enhanced visibility.
Importantly, such an obligation need not be an obligation contingent on whether any investigation is directed under section 36(4)(b), but must be an ongoing duty to obey obligations like prudential disclosures being enforced on critical infrastructure operators.
-
Unbundling Infrastructure and Ensuring Equal Access
However, transparency alone cannot neutralise exclusionary design when the algorithm itself functions as infrastructure. On platforms like Amazon and Flipkart, ranking algorithms do more than curate results; they embed operational preferences into the architecture of market access. Sellers who use the platform’s warehousing or delivery networks are rewarded with enhanced visibility, such as Buy Box eligibility and ‘fast delivery’ tags. Those who rely on independent logistics face algorithmic demotion in search rankings, often irrespective of product quality or consumer feedback. In this setup, the algorithm acts as a gatekeeper because it ties discoverability, which is the digital equivalent of shelf space, to a seller’s integration into the platform’s proprietary ecosystem. The result is not neutral optimisation but a structured form of exclusion, where access to consumers depends less on competition and more on conformity with platform infrastructure.
Although this behaviour cannot be described under the conventional vertical theories, it fully falls within the rationale of Article 5(7) of the DMA, which does not allow “gatekeepers” to impose requirements of accessing one core service on the use of another. To deal with this, India needs to follow a parallel measure by including an explanation to Section 3(3) of the Act, which states that linking fulfillment to visibility as a concerted practice occurs despite the lack of an express agreement. This would bring design-based bundling within the scope of per se scrutiny, treating it as horizontal exclusion between the platform and sellers not operating at the same level of trade.
-
Codifying Infrastructure and Onboarding Access Protocols
A further, subtler layer of distortion emerges in the context of selective onboarding, especially during high-stakes product launches. Here, platforms route early access to premium SKUs through a closed loop of preferred sellers, sidelining other capable vendors based on opaque and undisclosed criteria. While this does not amount to formal exclusivity, its effects are identical, as it prevents equal participation in commercially critical events and entrenches market segmentation.
To address this, India must codify Infrastructure and Onboarding Access Protocols under Section 64 of the Act. Drawing from Article 6(5) of the DMA and the UK’s SMS conduct requirements, these protocols should compel platforms to: (i) publish eligibility norms for warehousing and dark store access, (ii) disclose the extent to which infrastructure access affects ranking or promotional visibility, and (iii) establish objective and non-discriminatory criteria for onboarding into early-access or exclusive sales events.
Where such exclusion becomes systemic, the law must provide remedies beyond penalties. Section 27 of the Act should be amended to empower the CCI to impose structural remedies, including fulfillment neutrality and unbundling of platform-controlled infrastructure from algorithmic ranking layers.
Conclusion
The Amazon and Flipkart investigation reveals a structural transformation in how market power operates in the digital economy. Exclusion is no longer enforced through explicit contracts or overt collusion, but encoded in algorithms and infrastructure that shape access invisibly. These platforms no longer just host competition; they influence who gets to compete at all. By tying visibility to proprietary logistics, favouring select sellers through opaque ranking systems, and controlling onboarding into key sales events, they create an uneven playing field where access is determined by integration, not merit. To address this, competition law must evolve from focusing solely on anti-competitive outcomes to examining the architecture that produces them. Reclassifying such conduct under Section 3(3)(c) as horizontal market allocation correctly identifies the locus of harm in design-driven exclusion. However, retrospective enforcement is insufficient in digital markets where exclusion is fast-moving, opaque, and often irreversible. India must adopt a forward-looking framework that embeds algorithmic transparency, mandates fair access to fulfilment infrastructure, and codifies non-discriminatory onboarding protocols.
What is at stake is not just fairness to individual sellers but the integrity of digital markets themselves. When discoverability is no longer earned through quality or innovation but bought through infrastructural compliance, competition becomes a fiction. Restoring contestability requires more than policing bad outcomes; it requires setting rules for fair digital design. Ensuring that access to consumers is based on merit rather than manipulation is not just good regulation but essential for the future of India’s digital economy.