An e-commerce platform has two sides: buyers who procure and sellers who supply. Platform health depends on both sides being served. GeM has built formidable infrastructure for the buyer side: catalogue filtering, bid generation, demand aggregation, transaction fee architecture, MIS dashboards, payment integration. The seller side, in contrast, enters a helpdesk-and-escalation ecosystem with no published service-level commitments, no single point of accountability for incident resolution, no structured training infrastructure, and a VAE accommodation framework that relies on ad hoc engagement with GeM officials rather than codified process. How does GeM's seller-side servicing architecture compare with the way Singapore, Japan, China, and the United States have architected service for suppliers on their public procurement platforms, and what does the comparison reveal about the maturation arc still ahead for India's platform?
Every e-commerce platform has two sides. A marketplace without enough buyers starves its sellers of demand. A marketplace without enough sellers starves its buyers of choice. Platform stewards understand this symmetry; their investments in onboarding, retention, dispute resolution, training, and service velocity are distributed across both sides because the platform's long-term viability depends on both sides remaining engaged. Commercial marketplaces demonstrate this pattern in their architectures: seller central consoles with real-time support, dedicated account managers for high-volume sellers, structured seller education programmes, published resolution timelines for account restrictions, and dispute mechanisms with defined turnaround commitments. The reason is simple: a seller who cannot operate the platform, or cannot resolve a grievance within a reasonable window, leaves. When too many sellers leave, the buyer side begins to see a thinner catalogue, and the platform's value proposition erodes.
GeM has been, by a considerable margin, more attentive to the buyer side than to the seller side. The buyer-side infrastructure is sophisticated: catalogue filters that allow procurement officers to sort by supplier class, country of origin, and quality ratings; bid generation tools that translate specification documents into structured tenders; demand aggregation facilities that allow multiple buyer organisations to pool procurement for better rates; MIS dashboards that report budget utilisation, procurement cycle time, and category-wise spend patterns; training programmes administered through the National Institute of Financial Management and the National e-Governance Division; and an incident escalation architecture in which buyer complaints trigger immediate platform action. Every component of this infrastructure is calibrated for the experience of a procurement officer sitting in a Central or State government organisation.
The seller-side infrastructure, by comparison, is thinner. Registration is automated, which is useful. But once registered, a seller who encounters a problem, a product listing that has been rejected without clear reasons, a VAE application that is pending across quarters, an invoice that has been submitted but is blocked in a payment queue, a buyer-initiated incident that has frozen the seller's category listings, enters a helpdesk-and-escalation ecosystem whose service parameters are not published. There is no service-level commitment on VAE processing time. There is no published turnaround for incident revival. There is no single point of accountability for a stuck file; the seller must engage multiple GeM functionaries simultaneously, each of whom can advance or delay the matter independently. The structural accommodation for edge cases, such as the input-output invoice substitution for patented products, or the Annexure II documentation flexibility for foreign-manufactured items, emerges through one-to-one engagement with GeM officials in meetings; it is not codified, not reproducible for the next applicant, and not visible to RITES when it later reviews the file. A platform that has codified its buyer-side architecture through published rules, dashboards, and workflows has not performed the equivalent codification on the seller side. The seller's experience is still largely relationship-dependent.
The international comparison is instructive not because other platforms are perfect, but because the maturation arc they have traversed reveals what a fully architected servicing infrastructure looks like.
Singapore's GeBIZ, launched in June 2000 by the Defence Science and Technology Agency under Ministry of Finance funding, is a useful reference point because it is the closest structural analogue to GeM: a unified government e-procurement portal for the central state, mandated by policy, administered by a specialised agency. What distinguishes GeBIZ is the service-layer investment made explicitly for suppliers. DSTA collaborated with NUS Extension in 2004 and the Institute of Public Administration and Management in 2005 to deliver hands-on training to existing and potential suppliers, anticipating that a platform worth the investment required a supplier base equipped to use it. The Singapore Ministry of Finance publishes a formal Government Procurement Guide for Suppliers that documents the process end-to-end: how to register, how to identify opportunities, how to submit a bid, how to challenge an evaluation decision. When the platform was reviewed a decade and a half after launch, the redesign, delivered as GeBIZ 2.0, was explicitly framed around supplier user experience; the DSTA team conducted supplier workshops, interviews, and prototype testing cycles before finalising the architecture. The institutional disposition was that the platform's long-term health required active investment in the supplier side; it was not treated as an afterthought.
The United States operates a distinctly different architecture, but one that illustrates the depth of service infrastructure advanced procurement systems have found it necessary to build. The federal procurement platform is SAM.gov, the central registration system maintained by the General Services Administration. Around SAM.gov sits a network of over three hundred APEX Accelerators, formerly known as Procurement Technical Assistance Centers, administered by the Department of Defense and funded partly by the federal government and partly by state and non-profit partners. APEX Accelerators deliver free or low-cost one-on-one counselling to businesses pursuing federal, state, and local contracts. They help with SAM.gov registration, identify contract opportunities matched to the business's capabilities, advise on proposal preparation, and provide ongoing support through contract performance. Beyond APEX, the Small Business Administration's Office of Government Contracting deploys Procurement Center Representatives who review federal acquisition strategies before public announcement and advocate for small business set-asides, and Commercial Market Representatives who ensure that small businesses receive subcontracting opportunities through large federal prime contractors. The SBA also administers programmes such as the 8(a) Business Development Program, the Service-Disabled Veteran-Owned Small Business Program, and a series of small-business set-asides with a negotiated overall prime-contracting goal of 23 percent. The architecture is not simply a registration portal; it is a service ecosystem that the state has funded precisely because participation in federal procurement is operationally demanding and the state has decided that suppliers need institutional help to engage with it effectively.
Japan's architecture illustrates a different philosophy, service through institutional intermediation rather than direct counselling. The primary procurement notification vehicle remains the Official Gazette (Kanpo), and the Japan External Trade Organization maintains the comprehensive online database of procurement notices in English. JETRO's Business Support Centers across the country provide hands-on assistance to foreign suppliers navigating Japanese procurement, including interpretation of tender specifications and guidance on the Unified Supplier Qualification System that allows a single qualification to serve across all central government entities. The Ministry of Foreign Affairs hosts an annual Government Procurement Seminar where central government procuring entities walk foreign suppliers through their procurement plans for the fiscal year, a structured institutional opportunity to understand demand architecture before tenders are issued. For bid challenges, the Cabinet Office operates the Government Procurement Challenge System (CHANS), through which qualified suppliers can file complaints regarding GPA-covered procurements; the Government Procurement Review Board processes these complaints under a ten-working-day initial review window. The architecture distributes the service function across institutional intermediaries, but the function is unambiguously funded and staffed.
China's public procurement architecture is more fragmented, split between the Government Procurement Law administered by the Ministry of Finance and the Tendering and Bidding Law administered by the National Development and Reform Commission, with the ccgp.gov.cn portal as the centralised notification channel. The servicing infrastructure is less developed than in Singapore, Japan, or the United States, and foreign suppliers in particular face navigation challenges given the Chinese-language-only interface. However, the supplier-side architecture has seen consistent institutional investment: centralised procurement agencies have been established at provincial and central levels; the Ministry of Finance has issued procurement procedure regulations with defined timelines for bid evaluation and complaint resolution; the State Council has released successive action plans to strengthen procurement credit management, collaborative supervision, and platform upgrades. The 2025 Notice on Implementing Domestic Product Standards, effective January 2026, institutes a 20 percent evaluated-price advantage for products substantially transformed in China, but the associated compliance architecture explicitly specifies the declaration format suppliers must submit, prohibits procurement entities from demanding additional proof beyond the prescribed documents, and establishes disclosure requirements for supplier declarations. The architecture is imperfect and the supplier servicing infrastructure is not at the level of Singapore or the United States, but each procurement rule cycle has been accompanied by procedural specification that reduces the relationship dependency that characterises less mature platforms.
Read against these four comparators, GeM's position becomes clearer. India's platform is among the largest government procurement marketplaces in the world, second only to South Korea's KONEPS by Gross Merchandise Value, crossing Rs.13.60 lakh crore cumulatively by mid-2025. The scale of buyer-side activity is unambiguously successful. The institutional ambition, to bring multi-state, multi-category, multi-severity government procurement onto a single platform, has been delivered. What has not yet been built to the same level of maturity is the seller-side service infrastructure. There is no Indian equivalent to the APEX Accelerator network; there is no structured supplier-facing institution analogous to JETRO's Business Support Centers or Singapore's DSTA-NUS training collaboration; there is no formal equivalent of Japan's annual Government Procurement Seminar; there is no published service-level commitment analogous to Singapore's supplier guide or Japan's CHANS ten-working-day review window; there is no equivalent to the SBA Office of Government Contracting's small-business advocacy architecture, inside which Indian MSME suppliers, who collectively represent a substantial share of GeM's seller base, could find the kind of structured institutional support the federal US system routinely provides.
The gap is not a failure; it is a maturation stage. Every major procurement platform has traversed a similar arc. GeBIZ in its early years was a tendering-and-registration portal; the service infrastructure was retrofitted onto it as the platform's volume grew and as institutional learning surfaced the need. The US federal procurement system evolved its APEX Accelerator network and SBA service architecture over decades, each layer added as a response to observed supplier difficulty. Japan's Unified Supplier Qualification System and MOFA Government Procurement Seminars were institutional responses to the recognition that foreign suppliers could not effectively engage a Japanese-language-only procurement system without intermediation. The GeM platform has delivered the first generation of institutional architecture; the next generation, on current international comparisons, is the one that builds supplier service infrastructure with the same attention as the buyer-side workflows already demonstrate.
There are specific architectural choices that would produce immediate effect. A published service-level commitment for VAE processing, incident resolution, and account revival would convert the current relationship-dependent system into one where the seller knows what to expect. A single point of accountability for each seller case, functioning in the manner of a case officer rather than a distributed helpdesk, would eliminate the structural gap in escalation. A structured supplier training programme, potentially delivered through partnership with institutions such as the Indian Institutes of Management or the National Institute of Industrial Engineering, would create the equivalent of the DSTA-NUS collaboration that Singapore treated as foundational. An annual GeM Supplier Conference, at which Department of Commerce and the major buyer organisations walk the seller community through the forthcoming procurement pipeline, would replicate the institutional logic of Japan's MOFA seminar. An ombudsman function, analogous to the US Procurement Center Representative or Japan's CHANS Board, would give sellers a non-platform channel for grievances that do not respond to the standard incident management architecture. None of these requires statutory amendment. Each is an institutional choice about how the platform invests its service capacity.
The distinction between a procurement platform and a procurement marketplace rests here. A procurement platform matches buyers to sellers transactionally. A procurement marketplace sustains a community of buyers and sellers over decades. The first requires technology; the second requires service architecture. GeM has built the first with conspicuous success; the second is the unfinished work, and the international comparison shows what "finished" looks like. For Policy Access clients, the practical implication is that the service asymmetry on GeM is not a deficiency to be worked around; it is a structural feature that defines how the platform should be engaged today. For the institution itself, the service-side maturation is the next frontier, and every major procurement platform that preceded GeM made the investment because the platform's long-term institutional credibility depended on it.