How Indian D2C Brands Cut RTO From 39.2% to 21% in 4 Months

The number nobody puts next to CTR and ROAS

Every D2C dashboard in India tracks click-through rate, cost per acquisition, and conversion rate to two decimal places. Almost none of them give RTO — Return to Origin, a shipment that boomerangs back undelivered — the same scrutiny, even though it erases revenue after the sale is already booked as “won.” That gap is expensive, and it is now measurable at scale.

What changed, and when

Unicommerce’s India D2C Report 2026, drawn from 410 million-plus shipments across 6,000+ D2C brands on its Uniware platform, tracked RTO for optimised D2C sellers across the last festive-to-post-festive cycle. The number fell from 39.2% at the November 2025 festive peak to 25.6% in January 2026 to 21.0% by March 2026 — an 18-point drop, sustained over four months, not a one-week promo blip.

This is not a market-wide average. It describes brands that deliberately made three changes. The report is explicit that the fix was not a courier upgrade — these sellers didn’t switch logistics partners. They changed what happens at checkout and before dispatch.

The three fixes, in order of what they actually did

1. A prepaid incentive at checkout. Not removing COD — nudging the mix. COD orders in the same festive quarter returned at 58%, against under 15% for prepaid — roughly a 4x gap. An older Unicommerce index shows the same structural pattern further back: COD RTO at 20.9% versus prepaid at 5.8%, a 3.6x gap. This is not a festive anomaly. COD share is the single biggest lever inside any brand’s blended RTO number, which is why a ₹40–₹60 prepaid discount or free-shipping-on-prepaid offer moves the needle without touching a single delivery.

2. Pin-code-level courier routing. Instead of shipping every order through one contracted courier regardless of destination, route by actual delivery performance per pincode — Courier A clearing 97% in one pincode but only 78% in another is common, and routing around it is a data problem, not a relationship problem.

3. Address verification before dispatch. A confirmation step — SMS or WhatsApp — that catches an incomplete or wrong address before the shipment leaves the warehouse, not after it fails delivery three times and comes back.

Why this hits Tier 2/3 sellers hardest — and matters most there

Tier 2/3 cities now generate 66% of new D2C orders in India, a figure independently corroborated by a separate analysis of the same Unicommerce data. That is the growth engine of Indian D2C right now. It is also where address data is thinnest and COD reliance is highest — the same geography carrying the RTO liability is carrying the order growth. Cutting COD there to fix RTO would cut the growth. That is exactly why Unicommerce’s optimised brands incentivised prepaid instead of banning COD.

What an unfixed RTO number actually costs you

Cost-per-RTO figures vary by source and aren’t backed by one single audited study, but vendor estimates cluster in a consistent range: roughly ₹350–₹600 per returned order once you add forward shipping (₹50–120), reverse shipping (₹50–100), packaging (₹15–40), and the wasted customer-acquisition spend behind the original sale (₹150–500) — a worked example on a ₹1,200 AOV order puts the total near ₹561. Treat that as a directional estimate, not an invoice line.

For context on scale: an industry estimate places India’s annual COD-driven RTO losses above ₹20,000 crore — a figure attributed to RedSeer in secondary reporting that this research could not trace to a primary publication, so read it as a widely repeated industry estimate rather than an audited number. The point isn’t the precision of any single ₹ figure. It’s that the range is large enough to change how a founder prioritises Q4 budget.

The lever most sellers miss: how RTO actually gets triggered

RTO isn’t a courier deciding to give up. Standard practice among major Indian couriers — Delhivery, Xpressbees, Ekart, Bluedart — is three delivery attempts before a shipment is flipped to RTO. Every hour between “customer wasn’t available” (an NDR, or non-delivery report) and your response to it is an hour closer to that third failed attempt. Vendor estimates suggest compressing NDR response time to a same-day WhatsApp reschedule link recovers 35–45% of at-risk orders that would otherwise lapse into RTO — treat that as a directional benchmark, not a guarantee, but it costs process discipline, not capital.

Where your visuals fit into this

None of this replaces conversion work. It sits next to it. A listing photographed and styled to convert gets someone to click “buy” — but if a third of those buys boomerang back unpaid because of a checkout choice or a dispatch gap, the conversion-rate work is being quietly cancelled out by an operations gap nobody is grading with the same rigor as CTR. Both halves of that equation need an owner. If you don’t know your visual-conversion baseline against your RTO number in the same spreadsheet, that’s the first fix — a visual brand audit will tell you where the click is actually going, so the operational fix above isn’t paying for a leak upstream too.