RTO · Indian D2C
How to Reduce RTO Rates for Your D2C Brand in India
Return to Origin (RTO) is one of the most expensive problems in Indian D2C. You run ads, you generate orders, and then a chunk of them never actually reach the customer. They end up back in your warehouse after a round trip to the customer. Your ROAS looks healthy on paper, but your actual revenue has a hole in it.
The brutal part? Most brands are unknowingly making this worse through their ad campaigns.
Here's how BooleanMaths reduces RTOs and why fixing it requires more than just tightening your COD policies.
Why High ROAS Doesn't Mean High Revenue
Ad platforms like Meta and Google report conversions based on the purchase event, which they track on the thank-you page.
Meta sees the purchase event as the Final step in your marketing funnel.
But that purchase event fires whether the order gets delivered or returned. Your ad platform has no idea an order was cancelled at the doorstep three days later. So it keeps optimising toward users who place orders - including users who've already RTO'd multiple times.
From Meta's perspective, that repeat-RTO customer looks like a great customer who is purchasing multiple times.
This is how brands end up with a 4x ROAS and a 35% RTO rate at the same time.
The Three-Layer Approach BooleanMaths Uses to Cut RTOs
Method 1
Deep RTO Segmentation to understand who RTOs and why
Before you can reduce RTOs, you need to know where they're coming from. BooleanMaths gives you a breakdown across the dimensions that actually matter:
Payment method: COD orders RTO at significantly higher rates than prepaid. But within COD, there's more nuance: certain geographies, price points, and even order times skew the risk further.
Geography: Pincode and city-level RTO rates vary widely. Tier-2 and tier-3 cities are not monolithic; some outperform metro areas and some don't.
Behavioral signals: How a user browsed before placing an order, how quickly they checked out, whether they applied a discount code. All of these contribute to RTO prediction.
IP and device patterns: Repeat returners often exhibit consistent device or network signatures.
Delivery Timeline: RTO likelihood goes up as the delivery gets delayed. This might be because of a variety of factors
1. The customer had too long to change their mind
2. Or they simply forgot about the order
3. They ordered from multiple Brands and went with the first best option.
4. They has an urgent need and delayed delivery resulted in missing that window.
The output of this analysis is a clear picture of your high-risk segments, identifying specific clusters that you can act on.
Method 2
RTO Prediction at the Order Level
Knowing your historical RTO patterns is useful. Knowing the estimated RTO probability of an order as it comes in is better.
BooleanMaths uses the signals above to generate a risk score for each order at placement time. This gives your operations time to execute any of the following mitigation actions
Flag high-risk orders for IVR confirmation
Route the orders differently, or
Apply prepayment nudges before they ship.
Pre-emptively cancel for delayed shipments
This doesn't require overhauling your logistics, but making data driven decisions at each stage.
Method 3
Training Meta/Google Ad Algorithms on the right data signals
This is where most brands have a structural gap and where the compounding damage happens. Without enhanced tracking, your ad platforms are optimising on a low-quality purchase signal. BooleanMaths fixes this by sending enriched conversion signals through the Conversion API (CAPI), including:
COD vs. prepaid - prepaid orders are stronger delivery signals
Discounted vs. full-price - discount-driven orders have different fulfillment profiles
New vs. repeat customer
Bundle vs Single-Item purchase
High AOV vs Low AOV
Critically: Delivered vs Cancelled / RTO
That last one is the critical signal and it's not available when the order is placed. BooleanMaths integrates with your shipping partners, and the moment an order status is updated (delivered, RTO, cancelled), that event is sent to Meta and Google via CAPI.
Your ad platforms now knows the difference between a good order and a bad one. Over time, they stop optimising toward your high-RTO segments and start finding more buyers who actually convert to delivered revenue.
The Results
30%
Decrease in RTO rate within 3 weeks, with all three layers running in parallel
With segmentation insights, order-level prediction, and enriched CAPI signals running simultaneously, brands on BooleanMaths have seen RTO rates drop by 30% within three weeks. That's not a slow gradual improvement over a quarter, it's almost instant. Because the feedback loop to your ad platforms is real-time once enhanced signals are live.
The Bottom Line
If you're only looking at ROAS to evaluate your campaigns, you're looking at only half the picture. The other half is RTO. Every RTO order your ad platform helped acquire is revenue your business never saw and had to pay logistics cost on.
Fixing RTO isn't just an ops problem. It's also a data and attribution problem. And it starts with giving your ad platforms accurate signals about what a good order actually looks like.
