AI increasingly key to e-commerce players’ growth plans

For platforms like Meesho, investments in the technology are becoming significant growth levers; AI is also turning into a critical component of festive season sales
A person browsing an e-commerce app
Misrouting and failed deliveries are a major drag on e-commerce marginsVH/Unsplash
Published on: 

Leading e-commerce players are increasingly turning to AI and machine learning to improve everything from product recommendations to logistics and software development.

Take value-commerce leader Meesho. In its Q2 FY27 results, Meesho said its net merchandise value (NMV) grew 23% y-o-y to ₹12,911 crore. A major boost came from faster delivery times, which were about 15% lower year-on-year, helped by enhancements to its GeoIndia large language model (LLM), a denser logistics network, and better routing and partner allocation.

Misrouting and failed deliveries are a major drag on e-commerce margins.

While Indian players don't disclose numbers, according to fleet management company SmartRoutes, failed deliveries cost $17.2 per parcel in the US or £11.6 in the UK, with 70% of shoppers unlikely to return after a failure.

One way e-commerce platforms improve last-mile efficiency is by dividing postcodes into smaller delivery zones, or polygons, and allocating shipments based on delivery partners’ locations and availability, a former Flipkart executive who asked not to be named, told The Financial Standard. These zones can be adjusted to account for local traffic conditions and road closures, helping optimise delivery routes and reduce unnecessary travel.

In Meesho's case, GeoIndia solves a uniquely Indian problem: customers often don't write addresses in a consistent format, and landmark-heavy language such as "near the big banyan tree" or "behind Gandhi Market" are common, the company says. GeoIndia makes sense of such unstructured addresses to accurately identify customer location.

The efficiencies add up: Meesho’s contribution margin rose to 6% of NMV in the September quarter, from 4.6% in the previous quarter, said Vidit Aatrey, Meesho’s chairman, managing director, and chief executive officer. Lower logistics costs, -6.2% quarter-on-quarter per delivered order, helped improve the contribution margin.

AI is also speeding up software development. In its shareholders’ letter, the company said it moved code authorship to AI agents in the first quarter and made agents the default way of building software in the second quarter. Features that once took two to three weeks to develop can now ship in days, the company claimed. 

Despite the improvement in contribution margin, Meesho’s marketplace adjusted earnings before interest, taxes, depreciation and amortisation (EBITDA) loss stood at Rs 130 crore, or 1% of NMV, compared with 1.2% in the previous quarter. At work were higher server expenses, in addition to growing marketing costs, said Karan Taurani, executive vice president, Elara Securities.

E-commerce players also use AI to understand what sells where, who’s likely to buy what, Anand Ramanathan, partner and consumer industry leader, Deloitte India, told The Financial Standard. “And that pretty much decides where to set up their logistics and infrastructure.” It also helps companies decide how to get closer to existing consumers, optimise delivery and warehousing costs, and fulfil orders quickly in line with forecast demand, he said.

For instance, Amazon says it has been using AI to improve customer experience for over two decades. “The personalised recommendations customers get when they shop the Amazon.in marketplace, the pick paths in our fulfillment centres, and the conversational capabilities of Alexa are just a few examples of experiences fuelled by AI,” the company says. “And we believe generative AI is going to change virtually all customer experiences that we know.”

Amazon CEO Andy Jassy said last year that around 300 million customers used Rufus, its shopping assistant chatbot, and those users were about 60% more likely to complete a purchase. 

Investments in AI are becoming particularly vital for e-commerce companies during the festive sales season. According to InMobi Advertising’s "The Marketer’s Guide to India’s Festive Season 2026," 4 in 5 shoppers are actively seeking AI-powered shopping assistants to guide their festive purchases, and 3 in 5 plan to spend over ₹10,000 this festive season.  

AI is now actively reshaping how shoppers explore, evaluate, and decide on purchases, the report adds. From smart search suggestions (89%) to virtual product try-ons (81%), AI-assisted brand discovery and decision-making are gaining significant traction.

Meesho’s Attrey said the company will reinvest its efficiency gains into better consumer and seller experiences. It also aims to accelerate innovation and build a larger, more efficient marketplace that can generate sustainable free cash flow over the long term.