One-visit doorstep size swaps that turn fashion returns into exchanges
TL;DR Keepsize is a city delivery service for D2C fashion labels. When a shopper asks to return a garment for size, a Keepsize rider brings the other size, waits while it's tried on and takes back the first one in the same visit. Labels pay per swap and keep the sale.
Indian fashion returns are high, and most of them come down to size. A 2026 benchmark puts apparel returns at 25–40% of cash-on-delivery orders and 12–18% of prepaid ones, costing ₹200–280 each[1]. Back in 2019, the fashion head of one large marketplace said most apparel returns came from people wanting a different size[2].
A size return should be an easy save, since the shopper still wants the garment. In practice an exchange means one courier trip to collect, a check at the warehouse and then a second shipment days later, so plenty of shoppers give up and take the refund. The label pays for two or three legs and loses the sale anyway.
The seed proposed AR try-on to stop returns before they happen. I moved the fix to after the order. A try-on overlay shows how a kurta looks, not whether it fits at the shoulders, and size advice before purchase is already covered by another idea in this directory.
Here's how it works. Keepsize plugs into the label's returns page, and when a shopper picks "wrong size" the page offers a swap within 48 hours instead of a refund. Keepsize pulls the new size from a small stock pool it holds for each label in the city, or from the label's local store. The rider carries it to the door, the shopper tries it on and the rider checks the first piece for tags and damage before taking it back. If the new size also fails then the rider can process a normal return on the spot.
Returned pieces go back into the city stock pool, so the label doesn't ship them across the country and back.
The shopper pays nothing extra and the label pays per completed swap. The assumed price is ₹180–250 a swap, set below the ₹200–280 a standard return already costs[1], plus a monthly fee for holding stock. Your buyer is the head of operations at a D2C fashion label with steady volume in one or two metros.
Labels already use returns software to push exchanges and store credit over refunds, and one returns-software firm, for one, sells "smart exchange" and store-credit flows[3]. Software can offer an exchange but it can't make it fast. Dense metro delivery networks built for quick commerce have made same-day and next-day rider routes cheap enough to try this.
No public figure exists for exchange services in India. The useful number is the leak - at 25–40% COD returns and ₹200–280 per return[1], a label shipping 10,000 orders a month in one city may handle 1,500–3,000 returns, by rough arithmetic. If half are size-led and a third become swaps then that's 250–500 swaps a month per label in a single city.
Returns-software tools push exchanges in software but rely on ordinary courier pickups[3]. Fit tools work before purchase and claim fewer fit returns when shoppers follow advice[4]. Large marketplaces run their own logistics. We found no service that does a one-visit swap for independent D2C labels, though the search was limited.
| Difficulty | Moderate (39/100) |
| Starting capital | ₹80 L–1.5 Cr |
| Time to launch | 4–6 months |
The software is simple. The hard parts are running riders who handle garments carefully, holding the right sizes in each city without tying up a label's stock and keeping swap visits dense enough to make money. Labels will also want proof that riders don't damage garments or upset their customers.
No sector licence is needed for last-mile delivery of a client's goods, but people and stock bring their own rules.
Building this outside India? The categories above usually have a local equivalent, so check with your own regulator.
Don't trust AI to make the final call on damage disputes or refunds. A person should decide anything that costs a shopper or a label money.
Want to work out the AI layer for Keepsize? Talk to Yash →
Here's what can go south -
I like this for a founder who has run last-mile operations and wants to sell to fashion brands. If you're after a software business then look elsewhere. Prove two things first - that half of size-return requests at two labels accept a swap, and that swaps cost less than the returns they replace.
Plan for roughly ₹80 L–1.5 Cr over 4–6 months, mostly on riders, city stock rooms and insurance.
No sector licence. You need company and GST registration, DPDP compliance for shopper data and proper labour arrangements for riders.
Only with dense routes. Each swap has to cost less than the ₹200–280 a normal return already costs, so enough swaps per area decides everything.
The free Keepsize validation kit has the line-by-line capital breakdown, the effort scoring, the data and labour checklist, a 14-day validation sprint with pass/fail targets, a 90-day launch plan, 10 interview questions for labels and shoppers, and 5 assumptions to test. Download the kit →
Thinking about building Keepsize or something close to it? The make-or-break work is getting enough swaps per route that each visit costs less than a return, and that's a numbers problem I enjoy working through with founders. If you need a hand, you know where to find me. Talk to Yash →
For illustration only. Names are fictional and ideas hypothetical; sourced facts are real but may be out of date. Nothing here is legal, financial or professional advice, and Yash Khatri is not personally liable for any use of it. Read the full disclaimer.