The earnings call can be accessed here.
Meesho’s Q4FY26 earnings call reveals how the company deploys artificial intelligence (AI) across user acquisition, advertising, product discovery, logistics, and financial services. Here are five AI developments Meesho management disclosed on the call:
1. Vaani: voice AI to crack rural India: “We saw that with Vaani, we were able to convert more rural customers on our platform, and because of that, our customer acquisition cost (CAC) went down,” said chief executive officer (CEO) Vidit Aatrey.
Vaani is Meesho’s voice AI shopping agent. It lets users shop by speaking rather than typing, removing language and literacy barriers for rural, first-time internet users.
Explaining how Vaani reduces friction for new users, chief financial officer (CFO) Dhiresh Bansal said, “With the improvements and products that we have made through AI and otherwise, that barrier threshold has come down. Vaani, for instance, our voice agent, or other investments that we have done in terms of algorithms as well, have reduced that friction between a user installing the app and placing their first order.”
On why rural India still represents a large untapped opportunity, Aatrey said, “India has quite low in terms of number of people who are transacting online as a percentage of smartphone users. Even in other emerging markets, that number is north of 80%. India is closer to 30%.” Meesho closed FY26 with 264 million annual transacting users (ATU), up from 199 million in FY25.
On why going deeper into rural India has not hurt order frequency, Bansal said, “Almost a doubling of frequency has happened from a baseline perspective in the last three years. And that is a trend which is more than offsetting any potential socioeconomic impact that we might see when it comes to us going into deeper parts of India.”
Vaani featured in the Q3FY26 call but Q4 marks the first time that management credited it with a measurable reduction in CAC.
2. Algorithmic ad auctions: hook sellers first, charge later: “We have built sophisticated systems that now use AI to do the right targeting and get the right return on investment (ROI) for these sellers,” Aatrey said on the Q3 call, referring to how Meesho’s ad system works without asking sellers to pick keywords or audiences.
Most advertising platforms ask sellers to choose their own keywords or target audiences. Meesho does the opposite. Its ad system runs return on ad spend (ROAS)-backwards, meaning the algorithm decides which products to show which users based on what is most likely to result in a purchase. The seller sets a budget; Meesho’s AI decides the rest.
Q4 marks the first time management stated the two-phase pricing strategy explicitly. On phase one, referring to why Meesho keeps ROAS high right now, Aatrey said, “As soon as a lot of people are becoming advertisers for the first time on our platform and if they become advertisers and they get very good ROAS, the chances that they will spend more on the platform and put more budgets end up being much higher. So I think that’s the stage we are in.”
On phase two, referring to when Meesho plans to raise ad prices, he said, “At some point in time we’ll start to take pricing up to grow our revenue. We just want to be sure that we are doing it at the right time so that our steady state ad revenue is much higher than what we have today.”
On seller adoption, Aatrey said, “The number of catalogs or products which are now live on ads has improved significantly on a year-on-year basis (YoY), more than 40%. And that is the key input that eventually goes into a lot more competition in terms of ads and hence higher ad revenues for us.”
More than two thirds of sellers by gross merchandise value (GMV) now use the ad product, and seller ad budgets more than doubled in one year. Once ad prices rise, sellers who do not advertise will lose discoverability to those who do.
3. Relevance algorithms and content commerce: “We continue to make our price proposition better, our relevance algorithms better, and hence users are able to discover a lot better things in their first sort of year itself versus what they were doing earlier,” Aatrey said, referring to how AI improvements drive higher order frequency among new users.
Meesho’s feed personalises what each of the 264 million users sees based on browsing history, past purchases, location, and behaviour.
On how granular this personalisation is, Aatrey said, “The feed algorithms that we have, they are personalised for each and every user and hence the likely overlap between users kind of using a single sort of device to place orders through Meesho is relatively low.”
Content commerce, algorithm-driven social-style video and image feeds that trigger purchases, similar to TikTok Shop, sits on top of this.
Referring to content commerce as a new driver of first-year frequency, Aatrey said, “With propositions around Mall, content commerce, etcetera, also getting introduced and leading to more transactions, that also leads to higher frequency in the first year itself.”
Meesho launched its content commerce business in FY24 and introduced Creator Club in February 2025, a platform that lets creators collaborate directly with sellers and monetise content. Active creators reached 50,319 in the 12 months ended September 2025.
On how much headroom India has for frequency growth, comparing it to value-focused e-commerce players in China, Aatrey said, “If you look at frequencies of comparable sort of value-focused e-commerce companies in China, those would be closer to almost 100 times. So I think there is a lot of space to play.”
4. BNPL: AI-enabled credit for rural first-time buyers: “Going forward, as we kind of continue to invest behind BNPL and we’re seeing good sort of early signs there, that will become a driver,” said Bansal, referring to buy now pay later as a future lever for converting cash-on-delivery users to prepaid.
Buy Now Pay Later (BNPL) lets buyers receive goods immediately and pay in instalments later, without a credit card. For rural first-time buyers with no credit history, BNPL offers one of the few ways to access credit online. Meesho uses AI to assess creditworthiness for users with no formal financial history.
On why BNPL has not yet replaced other prepaid conversion methods, Bansal said, “BNPL is still fairly early in its life cycle. Right now the key initiatives that we have had within our prepaid kind of product, which is payment before delivery (PBD), the cost of prepaid orders kind of coming down and we passing consequentially some of them back to our consumers with better prepaid discounts have been the driving force so far.”
BNPL did not feature in the Q3 call. Q4 marks the first time management confirmed it as live and scaling.
5. Valmo logistics automation: “Will we bring a lot more automation within sort centres? I think the answer is yes, that’s a big focus area for us. We’re already experimenting with it in several places and our goal is in the next few years to basically bring state-of-the-art automation within all our sort centres,” said Aatrey, referring to Meesho’s plans to move Valmo beyond software into physical infrastructure.
Valmo is Meesho’s in-house logistics orchestration system. It uses software to decide which logistics partner handles which order, which route to take, and how to manage delivery outcomes.
In Q3, Valmo operated as an asset-light software layer. Q4 adds physical automation inside sort centres, the facilities that receive parcels, sort them by destination, and dispatch them to last-mile delivery partners.
On whether Meesho will own the automated sort centre infrastructure itself, Aatrey said, “We haven’t pinned down a specific model. We continue to kind of see pretty good encouraging results from some of our partners who would want to kind of invest and set up some of these facilities on our behalf.”
Valmo handles roughly 50% of Meesho’s order volumes. Automation will reduce per-order costs and affect the labour structure of the facilities.
Also read:
For You
此内容由惯性聚合(RSS阅读器)自动聚合整理，仅供阅读参考。 原文来自 — 版权归原作者所有。