During the 2025 holiday season, AI and AI agents helped generate 262 billion dollars in online retail sales worldwide, roughly a fifth of every online purchase made globally, according to Salesforce data covering more than one and a half billion shoppers. That is not a forecast for some future quarter. It already happened. However, most Indian D2C brands built no strategy around it because the term used to describe it, agentic commerce, still sounds like a buzzword rather than a business reality.
Every few years, a new term takes over marketing conversations before anyone has agreed on what it actually means. Agentic commerce has become the latest one. Founders use it in pitch decks, consultants use it in slide titles, and increasingly, it shows up in boardroom discussions, but very few know what exactly it means as a ground reality. It is worth pausing and asking a simpler question. What is actually happening, and why should a direct to consumer brand in India care right now rather than next year?
What is happening around the world
For two decades, a customer looking for, say, a pair of running shoes would open a search engine like Google or Bing, browse a few links, compare prices across tabs, and land on a brand’s website to buy.
That entire journey is now being compressed. Now, a shopper simply tells an AI assistant what they need, a running shoe under a certain budget, suited for trail use, deliverable by a certain date, and the assistant does the searching, comparing. Further, in a growing number of cases, the agent also buys with minimal human intervention. Another example would be the newer smart fridges, that assess if a family is low on milk, groceries, and orders these small ticket items based on past purchases. The AI agent is becoming the shopper’s proxy.
The world is already seeing the impact of this change. McKinsey‘s research from October 2025 estimated that AI agents could mediate between three trillion and five trillion dollars of global consumer commerce by 2030, with close to a trillion dollars of that concentrated in the United States retail market alone. Salesforce‘s own data from the 2025 holiday season, gathered from over one and a half billion shoppers globally, found that AI and AI agents already influenced twenty percent of all online retail sales during that period, worth 262 billion dollars. Retailers who had deployed their own AI shopping agents grew sales nearly sixty percent faster than those who had not.
Realising the shift in the way people intend to shop, the infrastructure to support this is being built at a pace that is unusual even by technology standards. In January this year, Google announced the Universal Commerce Protocol at the National Retail Federation’s annual conference in New York, an open standard co-developed with Shopify, Etsy, Wayfair, Target and Walmart, and endorsed by over twenty payment and retail companies including Visa, Mastercard and Stripe. It lets AI agents discover products, check availability and complete checkout inside Google’s AI Mode and Gemini, without the shopper ever leaving the conversation. OpenAI and Stripe had already released a similar open protocol months earlier, which now powers shopping features inside ChatGPT. These are not experimental features buried in a settings menu. They are becoming the plumbing of online retail.
Where India fits in
India is not watching this from the sidelines. In October 2025, the National Payments Corporation of India, along with Razorpay and OpenAI, launched a pilot allowing ChatGPT users in India to shop and pay through UPI without leaving the chat window, starting with BigBasket and Vodafone Idea as early merchant partners. By February this year, the same collaboration extended to Anthropic’s Claude, bringing Zomato, Swiggy and Zepto on board. What makes India’s case distinct is that UPI already gives the country real time, consent based payment rails that most markets simply do not have. Global platforms need to build new payment infrastructure for AI agents to transact safely. India built that infrastructure years ago for a different reason, and is now repurposing it for this one.
What this changes for D2C brands
For a D2C brand, all of this changes something fairly fundamental about how customers are won. A brand’s website, its search engine rankings, its paid advertising, all of that assumed a human being was doing the browsing and forming an opinion based on design, imagery and persuasive copy. An AI agent evaluating options on a shopper’s behalf does not respond to a beautifully shot product photograph or a clever tagline the way a person does. It reads structured data. It checks whether stock information is accurate, whether the return policy stated on the website matches what is in the product feed, whether pricing is consistent across channels. India’s own D2C sector, now valued at over a hundred billion dollars and still growing at a rapid pace according to Mordor Intelligence‘s market research, has spent the last decade getting extremely good at the human facing side of this equation, visual storytelling, influencer partnerships, festive campaigns. Very few brands have spent equal effort making sure their product data is something a machine can read cleanly and trust.
Being known by people and being legible to an AI agent are becoming two separate disciplines, and a brand can excel at one while being invisible on the other. A D2C brand with loyal customers and strong reviews could still lose a sale simply because its product feed had an inconsistency an agent flagged as untrustworthy, or, worse, because an AI agent could not read the website well enough to find those reviews at all.. Thus, investing in GEO and becoming ready for ACP is critical. The brands that move now, will always reap the benefits of first movers advantage.
Where Should A D2C Brand Invest
The first priority is to make the brand’s website machine-readable and trustworthy. That means clean, structured and consistent product data across the website and marketplaces, including pricing, inventory, delivery timelines, specifications, reviews and return policies. But trust will not come from a brand’s own website alone. Third-party validation, customer reviews, expert opinions, credible media mentions and consistent information across multiple sources will increasingly influence how AI systems assess whether a brand is worth recommending. GEO, therefore, is not just about being visible. It is about building enough credible signals for AI to understand and trust the brand.
The second priority is agentic commerce readiness. Brands should prepare their product catalogues, inventory and checkout infrastructure to work with emerging commerce protocols and AI agents. As AI systems become the new intermediaries between brands and consumers, brands with accurate first-party data and strong third-party trust signals will have an advantage. The objective is simple: when an AI agent evaluates multiple options, there should be enough reliable information across the web for it to confidently choose and recommend the brand.
Looking Forward
For Indian D2C brands, agentic commerce represents a shift in how discovery, consideration and transactions could happen. As AI increasingly becomes an intermediary between consumers and brands, factors such as machine-readable product information, first-party data and third-party trust signals will play a greater role in influencing which brands get discovered and recommended. The pace of adoption may vary, but the direction is clear. For brands, preparing for an AI-led commerce environment is no longer about chasing a new technology trend, but about being ready for where digital commerce is heading.
