Google Turns Gemini Into a Checkout Button for Flipkart in India

What actually happened this week

Google started a limited test in India where select users can buy products from Flipkart directly inside Gemini and AI Mode without leaving the chat interface. The rollout covers a small slice of inventory and users right now, with a broader push planned for later in October. Walmart owns Flipkart, so this is effectively Google letting a rival retailer's catalog live inside its AI assistant. We noticed the announcement buried in a TechCrunch piece on September 26, not a Google blog post, which tells you how experimental this still is.

Why this isn't just another shopping integration

Most shopping integrations redirect you to a merchant site. This one keeps you inside the LLM. That changes the economics of product discovery because Google now controls the entire funnel from intent to purchase. Technically, Gemini has to handle real-time inventory, pricing, payment authentication, and order confirmation without hallucinating a single SKU. We've seen Perplexity try this with Shopify, but Google has the payments rail (GPay), the identity layer, and the distribution advantage of being the default on Android. That combination makes this structurally different.

The second-order effect nobody's discussing

If this works, Google doesn't need to send traffic to Flipkart's app or website anymore. It becomes the storefront. That means Flipkart loses the direct relationship with the customer, the retargeting data, and the ability to cross-sell on its own terms. Walmart accepted this trade-off because distribution in India is brutal and Google owns the top of funnel. But we're watching whether Flipkart tries to claw back control by withholding its best inventory or pricing from the Gemini channel within six months.

How this compares to the Perplexity-Shopify deal

Perplexity launched a similar buy-in-chat feature with Shopify merchants in early 2025. It got decent press but almost no adoption because Perplexity lacks a native payments system and most users don't have it as a default interface. Google's version has GPay pre-loaded on every Android phone in India, plus the search habit. The lesson from Perplexity is that the tech works but distribution decides everything. Google is the only player that can make this a habit rather than a novelty.

What this breaks for traditional SEO and affiliate sites

Indian publishers and coupon sites have built entire businesses on ranking for product queries and earning affiliate fees from Flipkart and Amazon. If Gemini answers 'best phone under 20000 rupees' with a buy button, that traffic evaporates. We're already seeing click-through rates drop on commercial queries where AI Overviews appear. This test accelerates that by removing the click entirely. The affiliate model in India might not survive the next eighteen months if this scales.

The unspoken implication for Amazon India

Amazon isn't part of this test. They've spent a decade building their own assistant (Alexa) and their own shopping graph. But Alexa never became a shopping interface in India the way Google Search did. Now Google is effectively telling Amazon: your catalog stays on your site, but the buying intent we capture goes to whoever partners with us. Amazon's counter-move will likely be aggressive subsidy pricing on their app to train users to start there. Expect a price war on electronics and groceries by Diwali.

What we're watching for in the October rollout

Three signals will tell us if this is real or theater. First, does the catalog expand beyond electronics into fashion and grocery where Flipkart's margins are thinner? Second, does Google take a transaction fee or keep it free to lock in the behavior? Third, does the 'AI Mode' label disappear and this just become how Search works? If all three happen, we're looking at the first genuine post-search commerce loop at scale. If not, it's another Google experiment that gets deprecated by Q2 2027.

Why India is the only place this could launch first

India has 700 million smartphone users, UPI as a universal payments rail, and a population that skipped desktop entirely. The US has fragmented payments, higher customer acquisition costs, and entrenched Amazon Prime habits. Google tried buy-on-Google in the US years ago and killed it because merchants didn't adopt. In India, Flipkart needs Google more than Google needs Flipkart. That asymmetry lets Google dictate terms. We wouldn't bet on a US version until 2028 at earliest.