Amazon India, Flipkart optimize product listings for LLM chatbots

Amazon India and Flipkart are piloting product listing optimization for some categories to improve visibility on large language model (LLM) chatbots, according to people aware of the matter.

Amazon began its pilot after Diwali sales, while Flipkart is in talks with third-party firms that offer generative engine optimization services.

These moves follow the National Payments Corporation of India’s integration with ChatGPT in October, in partnership with Razorpay and Tata group-owned BigBasket, allowing users to shop from BigBasket and pay via UPI in a pilot.

Ecommerce product links on LLMs are ranked based on factors such as price, delivery date, inventory, geography, and description, said Mayank Mishra, founder of Consumable AI.

Experts say other ecommerce and quick commerce firms are likely to follow suit, as users increasingly use LLM search.

Amazon recently sent a notice to Perplexity, asking it to block its agent on the AI browser Comet from making purchases on Amazon on behalf of customers.

🔗 Source: The Economic Times

🧠 Food for thought

Implications, context, and why it matters.

Large language model (LLM) shopping in India is still early, with little data on transaction volumes

National Payments Corporation of India (NPCI) ran a ChatGPT with Unified Payments Interface (UPI) pilot in October with Razorpay and BigBasket 1. No public data says how many people finish buys through LLMs versus apps.
The integration is still a pilot. With no usage or conversion rate disclosures, it is tough to judge if tuning product listings for chatbots delivers Return on Investment (ROI).
Analysts see more ecommerce plus quick-commerce firms (instant-delivery grocers for essentials) following as users shift to LLM search. The size of LLM-initiated purchases is still unmeasured.
Amazon asked Perplexity (an AI search startup) to block its agent on Comet (an AI browser) from making purchases on Amazon for customers. Pushback on third-party AI commerce go-betweens could slow LLM shopping outside direct tie-ups.

Software-as-a-Service (SaaS) providers can build attribution and optimization tools for merchants working through LLM commerce

Consumable AI founder Mayank Mishra said LLM product rankings depend on price and delivery date. Inventory, geography, plus description affect placement. Merchants want tools to tune these inputs for chatbot visibility.
Software and marketing providers at third-party firms can build Generative Engine Optimization (GEO) services.
Investors could back platforms offering LLM-specific analytics and attribution, helping merchants measure which chatbot placements drive conversions versus visibility.
As integrations expand beyond ChatGPT to other AI models, multi-platform tools will be essential for merchants managing product listings across different AI systems.