What is ONDC (Open Network for Digital Commerce)?

The goal of ONDC is to create an inclusive e-commerce environment through an open protocol developed on open-source specifications, making it independent of any one platform.¹

Specifically, ONDC is a set of protocols and specifications that allow different e-commerce platforms to connect and interoperate with each other. This ability improves price discovery and provides consumers with more service providers to choose from.

ONDC represents a step toward digital commerce democratization, shifting it from a platform-centric model—where a few e-commerce giants dominate the market—to an open, interoperable platform where buyers and sellers can interact regardless of the platforms they’re using. It aims to move e-commerce away from walled gardens (closed, siloed, highly stratified ecosystems) to a more accessible environment that fosters competition, inclusivity and innovation.

ONDC was developed amid growing power concentration concerns in the e-commerce market during an era of rapid digital transformation. In India, as in many other countries, a small cohort of large e-commerce platforms grew into digital monopolies. This trend raised concerns about unfair trade practices, data monopolization and the marginalization of smaller, local businesses struggling to compete.

To address these challenges and democratize the space, the government of India (under the auspices of the Ministry of Commerce and Industry) began exploring the possibility of creating an open network for digital commerce. The goal was to build a framework that allows buyers and sellers to transact—regardless of the respective platforms they use—similar to how the Unified Payments Interface (UPI)² revolutionized digital payments by making them platform-agnostic.

In July 2021, the Department for Promotion of Industry and Internal Trade (DPIIT) set up a nine-member advisory council to steer the initiative and guide pilot programs in New Delhi NCR, Bengaluru, Mumbai, Bangalore, Coimbatore, Hyderabad, Pune and other cities. Within a few months, ONDC received its certification as a private sector-led, Section 8 nonprofit organization, after which several private sector organizations began integrating their apps and services with the ONDC network.

Since its launch, ONDC platforms have processed nearly 7 million orders, and the ONDC network has grown to more than 370,000 sellers and service providers across 588 “countable cities” (accounting for roughly 70% of Indian cities).³ ONDC protocols have been adopted by companies like Flipkart, Paytm, Pincode by PhonePe, Amazon, State Bank of India (SBI) and Meta (among others), with use cases that range from healthcare to fintech and agribusiness.

ONDC is also poised to become a game-changer in the food delivery sector, where it can drastically reduce commission rates compared to platforms like Zomato and Swiggy, which charge restaurants—many of which are small businesses—up to 25% commission.⁴

ONDC is not a single platform or app; it’s a series of protocols that enable disparate entities to transact, creating a level playing field for entrepreneurs, startups, small and medium enterprises, corporations and consumers.

The ONDC comprises three major components:

• Buyer apps: the applications that interact with demand-layer goods and services, such as chatbots, user-based apps and voice assistants.
  
  
• Seller apps: the applications that interact with supply-layer goods and services, such as receiving buyer requests, fulfilling orders, and publishing goods and services catalogs.
  
  
• ONDC gateways: the applications that orchestrate buyer and seller services. For instance, location-based buyer discovery apps like Magicpin.

ONDC is designed with a microservices architecture, meaning that each component of the network is an independent service that can communicate with other services through APIs (especially AsyncAPIs). This modular approach makes it easier for different platforms to integrate with ONDC. However, the open, accessible nature of ONDC requires certain IT components and processes to function optimally, hence the ONDC system’s reliance on the Beckn protocol.

The Beckn protocol—which is sponsored by the Beckn Foundation and owned by the Open Shared Mobility Foundation—is an open protocol that enables decentralized e-commerce and interoperability between online marketplaces, service providers and consumers.

Beckn gateways aggregate data and uncouple packet transmission layers from user experience layers so that transactions (such as order booking, payment, delivery and fulfillment) are standardized across the network. As such, they power the backend of the ONDC infrastructure.

The Beckn protocol includes:

As is the case with any complex system, maintaining the health and performance of ONDC platforms and services is essential. Application observability and API management solutions can be helpful in this regard, especially when used in concert with observability best practices. When deploying an ONDC infrastructure, organizations might consider:

This includes technical metrics (such as CPU usage, memory consumption and request latency) and business metrics (such as the number of transactions and success rates).

Logging practices include capturing detailed information (such as timestamps, error messages, stack traces and contextual data) about system behavior and errors. Each microservice, database and supporting service should also have its logs aggregated and stored in a centralized logging system for easy access and analysis.

Given the distributed, expansive nature of ONDC, distributed tracing—which tracks requests as they flow through various services—can prove invaluable to troubleshooting and optimization efforts.

Service-level indicators (SLIs) are the metrics that an organization measures for a particular service (such as uptime and error rate), while service-level objectives (SLOs) are the targets businesses set for these metrics (like 99.9% uptime). These measurements are defined and agreed upon in service-level agreements (SLAs). Defining relevant SLIs and SLOs can help teams establish expectations for system performance and availability, measure progress along the way and uphold client agreements.

Proactively testing the resilience of ONDC platforms by introducing chaos experiments (chaos engineering) enables teams to understand how the system reacts to unexpected failures and identify opportunities for improvement.

Artificial intelligence for IT operations (AIOps) can automate data correlation across monitoring tools, provide predictive system alerts and automate root cause analysis, all of which enhance ONDC observability.