Agentic Commerce - Exciting Stuff Ahead ? Or Here Already!

A couple of weeks ago at Google I/O, Vidhya Srinivasan, VP Advertising & Commerce at Google took the stage and introduced Universal Cart - an intelligent cart that works across Google – whether you're shopping on Search, Gemini app, YouTube or Gmail. It proactively spots price drops or when an item is back in stock, and uses intelligent reasoning to alert you about product incompatibility or payment method perk. This is on top of the infrastructure launched in the past including a common language for agents with https://ucp.dev/https://ucp.dev/ to the payments infrastructure to make agentic checkout seamless aka Agents Payment Protocol (AP2). Read more here.

A couple of months ago, I enjoyed Razorpay's demo on agentic commerce. Instead of navigating menus and making decisions along the way, users increasingly start with intent. They say what they want and expect the system to figure out the rest. Because, India doesn’t type. India talks. Find out more here.

https://www.youtube.com/watch?v=IBEckJgRpfE

And then, in September 2025, ChatGPT launched Instant Checkout - a feature launched by OpenAI on that let users purchase products directly within the ChatGPT interface without leaving the conversation. Users could describe what they needed, receive AI-curated product recommendations, and complete purchases with a single tap.

How it works: User asks ChatGPT for product recommendations > ChatGPT suggests options, lets you compare products > A "Buy" button appears on eligible listings> Payment is processed in-chat via Stripe using stored payment and shipping info. Built on the Agentic Commerce Protocol (ACP), an open standard co-developed with Stripe that enables AI agents, shoppers, and businesses to communicate for product discovery, checkout, and order fulfillment, at launch it supports Etsy sellers and Shopify merchants.

So there's a lot happening already in agentic commerce. And, it makes sense. It's a sure-shot way of making money. Driving consumption.

Defining Agentic Commerce

Agentic commerce is an emerging form of e-commerce in which autonomous AI agents independently execute purchasing and payment processes on behalf of users or organizations, combining advances in generative AI, autonomous agents, APIs, and digital payment infrastructures to direct transactions without direct human interaction. Unlike traditional e-commerce where users manually browse, select, and authorize payments at every step, agentic commerce transfers decision-making authority partially or fully to AI systems that can anticipate needs, navigate options, negotiate deals, and complete transactions through multistep chains of reasoning.

Why is this exciting for India?

Already, 84% customers use AI for shopping recommendations (Bloomberg). The Walmart Retail Rewired Report India 2025 found that 87% of Indian shoppers believe AI tools help them shop faster (Indian Express), while a Meta/IPSOS study found 80% of Indian festive shoppers used generative AI for gift research and inspiration. On the other hand, online shopping in India is highly fragmented across ecommerce, quick commerce and D2C websites. ONDC recognized this opportunity earlier, but never took off. Now the time seems just right.

Platforms like Swiggy, BigBasket, and Flipkart are already live with MCP implementations enabling AI-agent-driven ordering through ChatGPT and Gemini.

Problems to be Solve in Indian E-Commerce

Indian ecommerce is miles behind the US and China and faces a number of inefficiencies.

• E-Commerce Cart Abandonment

Cart abandonment represents the single largest source of value destruction, approximately 40% of total e-commerce leakage. Agentic commerce addresses this by completing transactions autonomously using pre-authorized payment protocols like UPI Reserve Pay, eliminating checkout friction entirely.

• Fragmented Inventory

A single brand operating across platforms like Blinkit, Zepto, Instamart, Flipkart Minutes, Amazon Now, plus FBA, FBF, Myntra-locked stock, and D2C channels can find itself financing 250+ distinct inventory pools, with every unit ageing differently. A bestseller can be unavailable in one pin code while identical units expire across town: lost revenue on one side, write-offs on the other with five billing formats, five return flows, and five reconciliation calendars replicated city after city. That's where a universal agentic commerce layer wins!

• Discovery Failures and Conversion Leakages

Nearly 80% discover products via social media but switch to separate commerce platforms to complete purchases, creating friction and drop-off. ~70% of urban Indian consumers use more than one channel before purchasing and poor product data infrastructure causes Rs 5,000 crore in annual losses through deficiencies in attributes, images, descriptions, and compliance disclosures that directly affect discoverability and conversion. Agentic commerce pools in information from the best sources available on the internet for that product, reducing the friction to cross-polinate information across distribution channels

Early Platform Adoption Trends in India

Flipkart: From Flippi to AI Storefronts

Flipkart's journey began with its ChatGPT-powered assistant "Flippi" in October 2023, evolving into "Shop Like a Pro" (SLAP) in January 2026 as a standalone conversational AI app. Flipkart is also among the founding partners of Google's Universal Commerce Protocol (UCP), an open standard enabling AI agents to complete shopping tasks across discovery, checkout, and post-purchase support within Google Search AI Mode and Gemini. The company is reportedly simultaneously building its own AI storefront to enable agentic commerce on platforms like ChatGPT, Perplexity, and Gemini.

BigBasket: First Indian Merchant on ChatGPT

BigBasket became among the first Indian merchants to enable shopping through ChatGPT, powered by Razorpay's payment stack and UPI infrastructure. Users can ask ChatGPT to order groceries. For example, "Help me order ingredients for a Thai-style vegetable curry for 4 people" and the AI agent checks the catalog, presents options, and places the order.

Amazon India: Alexa for Shopping

Amazon launched Rufus AI in India in August 2024, and then retired the brand in May 2026, unifying it with Alexa+ into "Alexa for Shopping," a personalized agentic assistant embedded directly in the search bar. The assistant can automate reordering of household staples, track prices, and build shopping carts based on stated preferences.

Meesho

Meesho launched "Vaani" in March 2026, a Gen-AI voice shopping assistant targeting tier II and III users, achieving 22% higher conversion rates and lower returns with over 1.5 million users in its first month. Notably, 73% of orders on Meesho do not begin with a search, relying instead on AI-driven feed personalization.

What it means for brands?

NIQ launched "Product Intelligence" in June 2026 to help brands transform fragmented product data into structured, AI-ready intelligence spanning 246 million unique items. Some brands already see AI agents driving approximately 10% of their revenue! An estimated 30% of generic D2C brands face obsolescence within 24 months if they cannot prove superior utility to AI agents

AI agents are replacing the traditional "search-scroll-compare" workflow with intent-driven, autonomous transactions, reducing retailers to "background utilities in agent-controlled marketplaces" . Until checkout occurs, the LLM owns all customer journey data and merchants cannot see how consumers arrived, why their brand surfaced, or what intent led to the purchase. AI agents operate on "cold calculus of utility optimisation" rather than brand affinity, reducing purchase decisions to functional specifications .

Razorpay's CPO states the opportunity for agentic commerce in India is "potentially as large as, if not larger than, UPI" .

Netcore's Agentic Predictions 2026 report projects autonomous AI agents will fundamentally reshape marketing, commerce, and growth accountability over the next 24 months. By 2029, 45% of routine enterprise tasks are expected to be handled by AI agents, with early adopters reporting productivity gains of 40–60%.

For brands and marketplaces in India, the strategic imperatives are clear: i) invest in machine-readable product data as the new brand moat, ii) build MCP implementations to remain discoverable by AI agents, iii) adopt UPI Reserve Pay for frictionless autonomous transactions, and iv) shift marketing budgets from attention-based advertising to utility-based optimization.

The shift from Attention Economy to Utility Economy means AI agents do not experience brand fatigue or respond to visual ads, forcing a pivot from creative-first to data-first business models

The Trust-Adoption Paradox

However, for all its excitement there are behavioural challenges. A LocalCircles nationwide survey (75k+ responses across 332 districts) found 73% of consumers concerned about personal data security, 69% citing limited transparency in how AI evaluates products, and 61% concerned about biased or sponsored recommendations. Trust deficits paradoxically also drive AI interest: 43% of online shoppers struggle to verify seller credentials and 40% question review authenticity, indicating AI adoption is linked to reducing search friction. The most telling indicator of this paradox: 88% of Indian consumers continue to choose cash on delivery despite AI adoption and 49% prefer AI tools embedded within existing e-commerce apps rather than standalone AI assistants (only 17% prefer external platforms). This suggests the path to agentic commerce in India runs through trusted platform ecosystems rather than standalone AI interfaces.

.