Nobody took your customer. What moved is the ten seconds in which they decide.
It used to be a Saturday evening on a laptop, or a phone on the sofa, with a bit of scrolling and a comparison or two. Now it’s 8:40pm, the app is already open because somebody is adding milk, and your category is a screen of nine tiles with a price, a pack size and a delivery time on each one.
That’s the whole change. The decision window collapsed from minutes to seconds, and it moved into an app that belongs to somebody else.
Which produces the question founders keep asking me in different words: should we be on quick commerce, and will it eat our website?
My answer, and I’ll defend it below. Quick commerce is not a sales channel you win. It’s rented shelf space, with the rent payable in advance, and it belongs in your plan as a trial and visibility channel while the repeat purchase stays somewhere you own.
Do not dismiss it, the numbers are real
Being contrarian about the scale would be silly.
Quick commerce crossed Rs 25,000 crore in GMV in 2025 and has been growing at over 40% year on year. Blinkit holds somewhere around 46% of it, Swiggy Instamart around 24%, Zepto around 22%.
For a founder in a snacking, beverage or personal care category, that is not a side experiment happening elsewhere. It is where a meaningful share of your category’s purchases now occur, and refusing to participate on principle is a decision to be absent from the moment of decision.
So the question is not whether to be there. It’s what you’re actually buying when you are.
Which categories are exposed, and which are not
The exposure is not uniform, and a lot of anxiety is misplaced.
Genuinely exposed: anything impulse-led or replenishment-led. Snacks, beverages, everyday personal care, over the counter health, pet food, supplements. Low consideration, familiar format, bought again without thinking.
Largely not exposed: anything that requires explanation, fit, or a decision somebody wants to feel good about. Higher priced skincare where the ingredient story does the selling, apparel, anything with a subscription logic, considered gifting, categories where your customer reads before buying.
If you’re in the second group, the quick commerce panic in your investor updates is borrowed from somebody else’s business. Check which group you’re in before you spend anything.
The economics, stated plainly
Here’s where I’d push back, and it’s a criticism of the fee structure rather than of any platform.
Published figures vary by category and by source, so treat these as ranges. Commissions across the three sit somewhere between roughly 8% and 25% depending on what you sell, with GST payable on the commission itself. On top of that come per unit inwarding and fulfilment charges, payment processing of a percent or two, and visibility spend that runs anywhere from 5% to 15% if you want to be seen at all.
Then the listing fee, which is the one founders underestimate. Blinkit’s is reported at around Rs 25,000 per SKU per state, returned as ad wallet credit that expires in twelve months.
Run that. Six SKUs across four states is roughly Rs 6 lakh committed before a single unit sells, with the refund arriving as a voucher you have to spend inside the same ecosystem.
And the arithmetic on a sale. A product carrying 60% gross margin at MRP gives up around a quarter of it to commission immediately, leaving something near 35% before you’ve paid for marketing or logistics. Reported totals put platform take at roughly 30% to 35% of revenue once fees, ads and operations are counted.
That’s a viable number for a brand with scale and turnover. For a self-funded brand it’s frequently the difference between a contribution margin and a hobby.
The cost that doesn’t appear on the invoice
Margin is the visible problem. The real one is that you don’t get the customer.
No name. No phone number. No order history you can act on. No way to tell a repeat buyer from a first timer, no way to send the second purchase reminder, no WhatsApp list at the end of it. The platform keeps all of that, because that is the actual product being sold to you.
Which matters more in India than it does in most markets, because the repeat conversation happens on WhatsApp and you cannot have it with somebody whose number you never had.
So the honest comparison is not 62% margin against 35%. It’s 62% margin plus a customer you can sell to again, against 35% and a stranger.
The part I’d actually worry about
Not cannibalisation. Your website was never winning the 8:40pm impulse purchase, so there’s less to cannibalise than the fear suggests.
What I’d worry about is funding your own commoditisation.
Go and look at your category screen in any of the three apps. Nine tiles. What’s visible on each is a photograph, a price, a pack size and a delivery estimate. Your ingredient story is not visible. Your founder’s reason for starting is not visible. Your packaging design is a thumbnail.
The only differentiators rendered on that screen are price and availability, and both of those are axes a larger competitor beats you on. Buying visibility to compete there is paying to be judged on the one dimension where your advantage doesn’t exist.
Every rupee you spend making that tile win is a rupee spent teaching your buyer that your category is interchangeable.
Distribution is worth paying for. Distribution that erases what makes you worth choosing is a different transaction, and it should be priced as one.
How I’d run both
Narrow, deliberate, and with the two channels doing different jobs.
Put two or three SKUs on quick commerce. Your best seller and the obvious impulse format. Not your catalogue.
Use different pack sizes across the two. A smaller trial pack on the app, the value pack and the bundle on your own site, so a price comparison does not produce a direct contradiction.
Treat the platform as paid trial with distribution attached, and judge it that way. What it should produce is people who have now tasted or used the thing. Whether it produces a second purchase from you, at your margin, is a question about what happens next.
Which is where the insert in the box matters. A card with a reason to come to your site, a code, a QR to a WhatsApp opt in. It converts modestly and it’s the only bridge you get from their customer to yours.
And fund it honestly. Quick commerce is not a cheap channel, and adding it while your existing acquisition is underfunded is the spreading-thin problem in a new costume. The marketplace-or-own-site arithmetic I went through when starting out applies here too, with a steeper take rate.
Where I come out
Use quick commerce for trial, presence and the impulse occasion you were never going to win otherwise. Keep repeat, margin and the customer relationship on ground you own.
And be clear-eyed that you are renting attention in somebody else’s store at a rent that rises when they need it to. That is a perfectly reasonable thing to do. It’s an unreasonable thing to build a brand on.
My notes on performance marketing cover the acquisition side. If you’d like to think through your own channel mix, drop me a line on email, WhatsApp or LinkedIn and we can have a quick chat. I’m contracted full time so this isn’t a pitch.