What Google's Algorithm Wants From Indian Brands in 2026 and How to Use It | Roma Datta Chobey, Google India

Your customer stopped searching in keywords. Your listings didn’t.

India’s e-commerce market is projected to reach $250 billion by 2030, up from around $90 billion today. Roma Datta Chobey, Managing Director of Connected Consumer Commerce at Google India, on what brands have to fix before they can claim any of it.

The Google and Deloitte report behind that number, The $250 Billion Commerce Frontier, identifies four forces expected to contribute roughly $100 billion of the growth: Inspired, Intelligent, Instant and Immersive. Behind them sits a demographic fact. A 220 million-strong Gen Z cohort is projected to account for 45% of online spend by 2030, as 150 million new shoppers come online and per capita spending doubles.

Roma spoke to Ragini about what that means at the level of an actual product page, and where the gap between consumer behaviour and brand behaviour is widest.

"Consumers are moving faster than businesses and brands. Brands are playing the catch-up game."

Searches are now three times longer

AI hasn’t replaced search so much as changed how people talk to it. Queries in AI Mode run almost three times longer than they used to, because people have stopped guessing at the right keyword and started describing the problem.

Her example: someone used to type "wireless headphones." Now they type something closer to "suggest the best headphones for a work from home setup, easy to store, easy to clean, cordless." The interface has become a kind of friend they explain themselves to. Visual search compounds it, with Google Lens letting a shopper point at a dress and ask for something similar in a pastel shade for a daytime event.

High quality information in, high quality answers out

The ranking logic hasn’t fundamentally changed; the language models sitting on top of it just make context matter more. Which puts an unglamorous thing at the centre of visibility: your product feed.

Sizes, current stock, whether the item is available in a particular store, and above all the product description. Roma’s point is that a description written for a keyword no longer does the job. Fabric, silhouette, occasion, who it is for, what weather it suits. If a shopper asks for a pastel dress for a daytime wedding, the listing that wins is the one carrying enough structured information for the system to recognise it belongs in that answer.

The customer doesn’t experience channels

Consumers move between YouTube, Search, a marketplace, an app and a physical store without registering any of it as a boundary. Roma is direct that brands have further to go here than consumers do, and recommends one unified strategy rather than separate online and offline ones: how data is stored, how marketing is designed, how logistics and operations run.

Her own example is mundane and exactly the point. She saw a dress online, checked whether the shop ten minutes from her house had it, went in to try two sizes, and looked at other things while there. The reverse happens just as often. Local inventory visibility is what converts a search into footfall, and the blocker, she thinks, is usually mindset rather than technology.

Creators have taken the place of your sister’s friends

Roma describes growing up shopping with her parents, then with her sister’s friends, and now largely online with creators occupying that same advisory role. The report projects creators will influence 30% of total retail spend by 2030, with one in ten online purchases linked to creator storefronts, particularly in Tier 2 and smaller markets.

The mistake brands make is treating this as a festive burst: hundreds of micro-creators, some free product, done. What works is matching the creator to the business objective. One brand targeting Tamil Nadu worked with ten trusted regional YouTube voices and saw an eightfold surge in brand searches. Values have to overlap and continuity matters, because a few messages here and there work for that period and then stop sticking.

Google’s position in this is deliberate. As Roma puts it, YouTube is not a marketplace, it is the matchmaker. Its Shopping Affiliate Program lets creators tag products so a viewer can click through and transact on the brand’s platform, with eligibility starting at 500 subscribers.

Virtual try-on stopped being a gimmick

The survey data undercuts the assumption that immersive tools are pitch-deck material. Of consumers who have tried virtual try-on, a meaningful share went on to buy, and that group spends more. Roma’s read is that the barrier is not interest, it is awareness and placement.

Buried three menus deep, a try-on influences nothing. Put it where the hesitation occurs, on the product page, in discovery, or visible on entering a store, and people use it. Consumers are willing. The friction is on the brand side.

Instant means something different in every category

Ten minutes matters for an ice cream craving at ten at night. For fashion needed tomorrow, instant means convenience rather than speed. Quick commerce is projected to reach $50 billion by 2030, with non-food categories driving 45% of that spend.

Beyond the metros, value and convenience both matter, and assortment turns out to be as much of a draw as delivery time. A local kirana carries trust but not variety. The report also identifies roughly 60 million consumers outside the metros who are digitally active but not yet transacting, which Roma frames as the clearest available opportunity. Reaching them means voice, video, camera search and regional languages, not a slower copy of the metro playbook.

Let the agent do the grunt work

Google’s framing on AI, she says, has been to be bold but be very responsible. On agentic commerce the objective is narrow: keep the joy of shopping for humans and give the grunt work to agents. She wants to plan a trip and choose the experiences; she does not want to monitor fare drops. The shopper sets the parameters and the range, and control stays with them.

The interoperability layer for this is Google’s universal commerce protocol, an open standard so that agents and platforms can speak a common language, built with a small set of design partners including Flipkart in India.

For a founder, the action is less futuristic. Ask your CTO about the quality of your feeds. How is product data structured, where are the silos, are the pipes ready. An agent cannot recommend what it cannot read.

Listen to the full conversation with Roma Datta Chobey on FyndOutWithRagini.