A customer places a cash-on-delivery order. Before the parcel ever reaches their doorstep, the brand has already paid Meta to acquire that customer, packed the product, handed it to a courier, and moved inventory out of the warehouse.

Then the customer refuses delivery.

The sale disappears. The costs don't.

No card was ever charged, so there was nothing to lose by saying no at the door. The parcel travels all the way back. The brand eats the ad spend, the packaging, the shipping both ways, and the shelf space that product could have used for something that actually sold.

Now ask the obvious question nobody in Indian D2C likes sitting with: what if that order should never have shipped in the first place?

That is the problem ERRA, founded by Ishmit Mehta, is trying to solve.

Cash on delivery exists in India for a simple reason: trust. Most shoppers buying from a D2C brand they have never heard of are not willing to pay upfront for a product they cannot touch or verify. COD removes that hesitation and is a major reason India's D2C boom was possible at all.

Ironically, the very payment method that helped India's D2C ecosystem grow has also become one of its biggest sources of profit leakage.

But COD has a side effect that rarely makes it into the pitch deck: Return to Origin, or RTO. This is what happens when a shipped order comes back unsold, whether because the customer refused it, gave a wrong address, changed their mind, or simply stopped responding.

The scale of this is bigger than most founders admit out loud. According to data compiled by Vyora, a platform focused on cutting COD returns for Indian D2C brands, the national RTO rate on COD orders runs at roughly 26 percent, against under 2 percent for prepaid orders, making COD returns more than ten times worse. In fashion and footwear specifically, RTO rates can climb as high as 40 percent. And while COD accounts for around 62 percent of order volume industry-wide, it is responsible for an outsized 76 to 83 percent of total RTO volume.

Most founders have quietly accepted this as the cost of doing business in India. ERRA is built on the bet that it does not have to be.

RTO is not a single line item. It is a chain reaction that touches almost every function in a D2C business.

The Meta or Google ad spend used to acquire that customer is gone regardless of whether the order is delivered. Packaging material is wasted. Warehouse labor is spent twice, once to pack and once to receive the return. The forward and reverse courier legs are both paid for. Customer support has to handle the confirmation calls and complaints. And the returned inventory sits idle, tying up working capital that could have gone into stock that actually sells.

One failed COD order is rarely just one loss. It is five or six small losses stacked on top of each other, spread across departments that rarely talk to one another about it.

To understand why this keeps happening at scale, it helps to look at the refusal itself, not just its cost.

Some refusals are impulse purchases the customer regretted before the parcel even left the warehouse. In some cases, brands also report fake or malicious orders placed with no real intent to buy. Some customers simply find the same or a cheaper product elsewhere while waiting for delivery. Delivery delays cause others to lose interest entirely. Wrong or incomplete addresses mean couriers cannot reach the right door in the first place. Occasionally someone other than the account holder placed the order without their knowledge. And a large chunk of refusals come down to something as mundane as the customer missing the delivery call and never following up.

None of these reasons are exotic. That is exactly the point. RTO is not one hard problem to solve, it is seven small, boring ones that happen thousands of times a day across the industry.

Before writing a single line of product strategy, Ishmit Mehta had already lived the problem she is now building a business around.

According to ERRA's own site, the product was not built by consultants or logistics middlemen. It came from someone who personally ran COD operations and absorbed the losses that come with them: shipping hundreds of COD orders herself, spending real ad budgets on orders that came back, dealing with fake delivery attempts with no way to prove they happened, and fighting non-delivery report disputes with no recording to back her case up.

As she puts it on the ERRA site, the company "exists because I paid for RTO failures I didn't create."

That line does a lot of work. It is the difference between a founder who read a market report about RTO and a founder who watched it eat her own margins in real time.

Most brands don't ignore RTO. They simply fight it with people instead of software.

Founders dealing with RTO today are not starting from zero. The common playbook includes manual calling teams that ring customers to confirm COD orders before dispatch, WhatsApp confirmation messages, and dependence on courier partners' own fraud and NDR handling.

The problem is that none of this scales cleanly. Manual calling is slow and expensive to staff. Confirmation windows matter enormously. GoKwik, which runs RTO intelligence across a network of roughly 180 million Indian shoppers, has found that response rates on COD confirmation calls drop by 40 percent just thirty minutes after checkout, yet most manual calling flows only reach the customer the next morning, by which point intent has already cooled. Courier-side fraud checks are also outside a brand's control, which is exactly the dependency ERRA is positioning itself against.

ERRA's public positioning is still early and light on technical detail, which tracks with its current stage. What is clear from the site is the intent: move RTO management from something a brand outsources to people and couriers, into a system the brand itself controls. The broad shape of the workflow is order placed, order verified, customer confirms, and only verified orders move to courier pickup, cutting down the number of doomed parcels that ever leave the warehouse. The specific mechanics of how verification happens have not been made public yet.

ERRA is currently in a founding cohort stage rather than general availability. The site is onboarding a small group of COD-first D2C brands shipping at least 50 orders a day that are actively losing margin to RTO or fake delivery attempts, offering a free pilot with no credit card required and founding pricing locked in for twelve months. Full pricing, and how it scales once the pilot phase ends, has not been disclosed publicly.

The logic of who pays and why is straightforward even without public numbers: any brand losing a meaningful percentage of revenue to RTO has a very calculable ROI case for a tool that meaningfully cuts that number, which is the kind of expense a SaaS subscription is easy to justify against.

The interesting thing about RTO is that nobody is trying to solve it in exactly the same way.

ERRA is not entering an empty market. GoKwik, one of the larger checkout and COD players in Indian D2C, launched a Shopify app called Kwik COD that draws on a large shopper network to flag risky orders, verify COD purchases, and offer partial-COD alternatives. Pragma runs a predictive RTO Suite that uses machine learning and pin-code level data to flag high-risk COD orders before they ship, paired with automated WhatsApp and SMS confirmations. Shopify's ecosystem also has smaller apps like COD Toolkit aimed at reducing fake COD orders through OTP verification.

So why would ERRA exist if this problem already has vendors attacking it? The honest answer is that the space is large enough, and the pain acute enough, that it supports more than one player, and ERRA's pitch leans hard on founder credibility over feature lists. Whether that is enough to win founder trust against more established, better-funded competitors is the real open question.

Three shifts make this a good moment for a company like ERRA to exist. Customer acquisition costs across Meta and Google have kept climbing for years, which means every RTO-lost order is now a more expensive mistake than it used to be. Shipping and reverse logistics costs have risen alongside it. And the broader mood in Indian D2C has shifted from growth-at-any-cost to profitability, which means founders who once shrugged off RTO as a rounding error are now under real pressure to fix it.

RTO used to be an accepted tax on doing COD business in India. It is increasingly being treated as a solvable inefficiency instead.

It's worth asking why a company like ERRA couldn't have existed five years ago. Part of the answer is that voice and verification technology has only recently gotten good and cheap enough to run at consumer scale. Part of it is that SaaS adoption among Indian D2C brands, even small ones, has become far more normal than it was in 2019 or 2020. And part of it is cultural: the D2C founders raising money and running brands today are simply less willing to treat growth as the only metric that matters, which makes a tool that protects margin a much easier sell than it would have been during the growth-at-all-costs years.

A few honest questions are worth sitting with before treating ERRA as a sure thing.

Can automated verification actually move the RTO needle by a meaningful margin, or does it mostly catch the easy cases while the harder ones, like genuine impulse-buyer regret, remain unsolved? Will D2C brands, many of which are small teams already stretched thin, trust a new automated layer in their order flow enough to adopt it at scale? Is there anything stopping a large logistics or checkout platform from building the same capability in-house and bundling it for free? And with pricing and detailed product mechanics still undisclosed, it is genuinely too early to say how defensible ERRA's approach is against better-capitalized incumbents.

None of these are reasons to dismiss ERRA. They are reasons to watch how the founding cohort's actual results play out before drawing conclusions.

The most interesting startups do not always invent a new market. Sometimes they just refuse to accept a cost everyone else has decided is unavoidable. RTO has been treated as background noise in Indian D2C for years, a tax nobody enjoys paying but nobody expects to eliminate either.

ERRA's bet is that this acceptance is the actual market gap, not a lack of technology. Every successful software company starts with a belief that an expensive habit isn't inevitable. ERRA is making that bet on RTO. The next 12 to 18 months will reveal whether Indian D2C founders agree.

This piece is part of The Entrepreneur Post's Business Decode series, breaking down how Indian startups are built.

Disclaimer: This article is for informational purposes only and does not constitute investment or business advice. TEP has no financial relationship with ERRA.