The Insight Labs Daily

★ Lead Story

Yesterday · 2 min read

Amazon sold 49% more groceries in India and lost three times as much money

Amazon Retail India, the entity that runs Amazon Fresh and the ten-minute service Amazon Now, filed its FY26 numbers this week. Revenue reached Rs 3,065 crore, up 49% on the year. Loss reached Rs 1,158 crore, up about 194%.

That is the sentence worth sitting with. Sales grew by half. Losses grew by two and a half times.

Amazon Now launched into a market that Blinkit, Instamart and Zepto had already spent four years and several thousand crore teaching Indian shoppers to expect. Entering late means buying attention that the incumbents already own, and buying it in the same three cities where their dark stores are densest.

The gap between the two growth rates is the price of that entry. Every additional rupee of revenue in FY26 came with roughly 38 paise of loss attached, against something closer to 19 paise the year before.

The question for the next filing is whether that ratio narrows as stores mature, or whether it is simply what third place costs.

The comparison that matters is not Amazon against Blinkit. It is Amazon against its own earlier grocery attempts. Amazon Fresh has been operating in India since 2020 on a scheduled-delivery model with warehouse economics. Amazon Now put a ten-minute promise on top of that, which means smaller stores, more of them, and a delivery fleet sized for the worst hour of the day rather than the average one.

Quick commerce rewards density in a very specific way: a dark store only works when the orders per store per day cross a threshold, and that threshold rises with rent. A late entrant has to open the store before the orders exist, and carry the empty months. Blinkit and Instamart carried those months in 2022 and 2023, when land and rider costs were lower. Amazon is carrying them now.

The caveat is that Amazon has a balance sheet that does not care. The parent has funded Indian losses for a decade without flinching. What this filing changes is not whether Amazon can afford to compete, but what the shareholders of the listed competitors can now read about the cost of holding them off.

Today's Top 5

5 stories

Oziva · Yesterday

A nutrition brand spent 53 paise of every revenue rupee on marketing

Oziva, the plant-based nutrition brand Hindustan Unilever took majority control of in 2022, grew FY26 revenue 80% to Rs 463 crore. Its marketing spend for the year was Rs 246 crore.

Eighty per cent growth is real growth. It was also bought, and the receipt is visible in the same filing.

For a brand inside HUL, that number carries an extra meaning. HUL's own marketing-to-sales ratio sits in the low teens. Oziva is running at four times that, which is what a category still being explained to the consumer costs.

Direct-to-consumer nutrition in India has a structural problem that no amount of scale removes: repeat rates. A shampoo buyer returns on a schedule set by the bottle. A protein or hair-gummy buyer returns only while they still believe the product is working. When belief lapses, the brand pays to reacquire the same person.

That is why the marketing line grows with revenue rather than flattening against it. The interesting number in next year's filing will not be revenue. It will be whether Rs 246 crore becomes Rs 300 crore on Rs 700 crore of sales, or stays flat while sales rise.

Parag Milk Foods · This week

Parag is putting Rs 100 crore into paneer, and quadrupling the line

Parag Milk Foods has approved Rs 100 crore to take its paneer capacity to 80 tonnes a day, roughly four times where it sits today. The stock rose about 2% on the announcement.

Paneer is the least glamorous line in a dairy portfolio and quietly one of the best. It carries a higher margin than liquid milk, needs no marketing to explain itself, and its demand curve follows the festive and wedding calendar rather than advertising.

Quadrupling a line is not a demand forecast. It is a bet that organised paneer keeps taking share from the loose block sold at the local dairy.

The share shift is the whole thesis. Most paneer in India is still bought unbranded, cut from a block, with no date on it and no cold chain behind it. Every year that packaged paneer gets closer in price, a slice of that volume moves across, and it does not move back.

The risk sits in the input. Paneer needs a lot of milk per kilogram, and milk procurement prices have been climbing. A four-times capacity commitment made at today's milk price looks different if procurement runs ahead of what the retail shelf will take.

Mahindra · 2 days ago

Mahindra has 50 public chargers and wants 250 sites by end of 2027

Mahindra's Charge_iN public charging network has crossed 50 stations, and the company has put a number on the next phase: 250 sites by December 2027. The first joint site with HPCL, a 180 kW fast charger, has opened on the Mumbai-Pune Expressway.

The HPCL detail is the one that matters. A carmaker building chargers alone has to find land, power and footfall. A fuel retailer already has all three, sitting idle between petrol customers.

Fifty stations is not a network. It is a proof that the partnership model works before the expensive part begins.

Indian EV buyers do not actually need highway chargers for daily use, and Mahindra knows it. Almost all charging happens at home overnight. What the highway charger removes is the reason not to buy, which is the single long trip a year that a buyer imagines before signing.

That makes charging a marketing cost as much as an infrastructure one, and it explains the site-count target. The number that would tell you whether the network is working is utilisation per charger, and nobody publishes it.

Google · 2 days ago

93% of Indian car buyers used an AI tool before they bought

A Google and Kantar study of 1,022 recent Indian car buyers found 93% used AI surfaces somewhere in their research, and 90% ran into an AI-generated summary while searching.

Car buying in India used to run through a showroom, a brochure and a cousin who knows about cars. The brochure went first, then the cousin got replaced by YouTube. The summary at the top of the search page is now taking the first pass.

For an automaker, that changes what the marketing budget is buying. A comparison written by a model is assembled from whatever the web already says about the car, and no amount of campaign spend edits it directly.

There is a second-order effect that brands have not priced yet. When a summary answers the question, the buyer never lands on the brand's own configurator, where the upsell lives. Reach can hold steady while the path that monetised it quietly disappears.

The defensive move is unglamorous: make sure the specification pages, service costs and ownership reviews that the summaries are built from are accurate and complete. That is a content operation, not a campaign, and most Indian auto marketing teams are not staffed for it.

Campus Activewear · 2 days ago

Campus redesigned its store and wants 60 of them in one region

Campus Activewear has opened its first reimagined retail format in New Delhi and set a target of more than 60 outlets across the NCR.

Most Indian footwear brands announcing stores this year are announcing geography, with a first shop in a smaller city. Campus is announcing density instead, in the region it already knows best.

Sixty stores in one metro cluster is a distribution decision dressed as a retail one. It puts the brand inside walking distance of the buyer at the moment the online price stops being the deciding factor.

Density has an economics that geography does not. Stores in one cluster share a warehouse, share staff during peaks, and let a brand run a same-day exchange that an isolated store in another state cannot. It also lets marketing spend work on a single media market rather than being spread thin.

The exposure is equally concentrated. A brand with 60 stores in NCR has bet its offline P&L on one region's rents and one region's festive season.

⚡ 30-Second Scan

Maruti Suzuki has commissioned a 300 kW green hydrogen pilot plant at Manesar, running on surplus solar from the plant's own rooftop rather than grid power.

Kinetic Engineering is putting Rs 57 crore behind exports and its EV subsidiary, against an order book the company puts at around Rs 500 crore.

Adidas has opened its first OCS-format store in Srinagar, taking the format into a market most global sportswear brands still serve through distributors.

Sourced from public reporting; analysis by The Insight Labs.

Sources: Entrackr, Business Standard, Autocar Professional, Storyboard18, Social Samosa, Indian Retailer.

Want the Tuesday deep-dive?

The Insight Labs newsletter · every Tuesday · one full FMCG case-study from inside the P&L. Free.

Subscribe →

Add Insight Labs to your home screen

The Insight Labs Daily.