The RTO Problem Nobody Prices In

Somewhere between fifteen and thirty percent. That's the range most Indian D2C brands live in for RTO, return to origin, and if that number surprised you, it's because almost nobody talks about it in public the way they talk about revenue.

RTO is what happens when a customer orders something, the courier ships it, and the customer doesn't accept delivery. The product comes back. You've now paid for logistics twice, packaging twice, and the order fulfilled zero revenue while consuming real cash both ways.

I ran a D2C fragrance brand that hit close to ₹55 lakh in monthly revenue and roughly 10,000 orders a month. RTO was one of the quiet taxes that never showed up on the headline slide, and it was one of the specific things I wish I'd priced in from month one instead of month fourteen.

Why customers don't accept delivery

The obvious answer is cash on delivery, and it's a real part of the story. When a customer hasn't paid yet, the psychological cost of rejecting a delivery is close to zero. They ordered on impulse, the mood passed, the courier showed up on a bad day, and refusing costs them nothing.

But COD is only part of it. Customers also reject deliveries because the product took longer to arrive than expected and they'd moved on. Because they found the same thing cheaper somewhere else in the meantime. Because the courier called at a moment they couldn't answer, and nobody rescheduled properly. Because the order was a considered purchase that got reconsidered the moment it was time to actually pay for it, even digitally.

None of these reasons are irrational from the customer's side. They're completely rational. The system simply makes rejection free for them and expensive for you, and rational actors respond to incentives, every time.

What RTO actually costs, beyond the obvious

The forward shipping cost is the easy part to count. You already priced that in. The reverse shipping cost is the part that quietly doubles the damage, because bringing a rejected package back through the courier network costs almost as much as sending it out.

Then there's the packaging, which in most categories can't be reused once it's been through two legs of transit and sat in a returns warehouse for a week. There's the restocking labour, someone has to receive it, inspect it, and decide whether it's sellable again or damaged goods. There's the inventory that's now tied up in transit instead of available for a customer who would have actually accepted delivery. And there's the marketing spend that acquired that customer in the first place, which you paid regardless of whether the order completed.

Add all of that up across a meaningful RTO percentage, and you get a number that looks less like a shipping line item and more like a second cost of goods sold, sitting quietly underneath the first one, that most P&Ls don't have a dedicated line for.

The category-by-category reality

RTO isn't a flat tax. It varies enormously by what you're selling, and pretending otherwise is one of the more common mistakes I've watched founders make.

Fashion and apparel run high, because size and fit uncertainty is baked into the category, and customers order multiple sizes with the explicit intent of returning what doesn't fit. Fragrance and beauty run moderate, driven mostly by impulse ordering and COD psychology rather than product uncertainty. Electronics and higher-ticket items generally run lower, because the customer has done more consideration before committing, and the absolute value at risk makes both sides more careful.

If your RTO rate looks unusually good or unusually bad compared to your category's typical range, that's worth investigating specifically, because it's telling you something about either your product-market fit or your operational process that the aggregate number is hiding.

The fixes that actually move the number, and the ones that don't

Prepaid discounts move the number, because they shift the psychological cost of rejection back onto the customer. If money has already left their account, walking away from a delivery now costs them something, and rational actors respond to that too.

Address and phone verification at checkout moves the number, because a meaningful share of RTO isn't rejection at all, it's an undeliverable address or an unreachable customer, which looks identical in your dashboard but has a completely different fix.

Realistic delivery timelines move the number, because the gap between what you promised and what you delivered is where impulse decays into indifference. A customer who expected two days and got seven has had five extra days to stop wanting the thing.

What doesn't move the number, in my experience, is aggressive customer support scripts trying to convince someone to accept a delivery they've already mentally rejected. By the time the courier is at the door, the decision was usually made days earlier, for reasons a phone call at the doorstep can't undo.

Why this connects directly back to contribution margin

RTO isn't a separate problem from unit economics. It's one of the specific inputs that decides whether your contribution margin is real or imaginary. A founder who calculates contribution margin using only completed, accepted orders and ignores the cost of the rejected ones is calculating a number that describes a business that doesn't exist.

The order that comes back isn't a neutral event. It's a negative one, and if your model treats it as simply absent from revenue rather than present as a cost, your model is lying to you in a specific and correctable way.

The honest version of the calculation includes RTO as a line item against every order attempted, not just every order completed, because that's the actual economic reality your cash flow experiences every single day.

The decision most founders avoid making

Here's the uncomfortable part. Sometimes the right fix for RTO isn't operational. It's structural. It's recognising that a specific channel, a specific price point, or a specific geography has a rejection rate that no amount of process improvement will meaningfully fix, because the underlying customer behaviour driving it isn't a bug, it's the honest response to how that channel and price point are being sold.

That can mean turning off cash on delivery for certain order values, even knowing it will reduce your top-line order count. It can mean deprioritising a marketing channel that brings in high-intent-looking traffic that converts to orders but not to accepted deliveries. Both of those decisions feel like giving up growth. What they actually are is giving up the specific kind of growth that was never real growth to begin with, just revenue on its way to becoming a return.

RTO is not a cost of doing business in Indian ecommerce that you simply absorb and move past. It's a number with causes, and every cause has a lever. The founders who treat it as weather, unpredictable and unavoidable, are the ones who keep paying full price for it, month after month, while the founders who treat it as a solvable input usually find that it was more solvable than the shipping company ever let on.