COD-adjusted Break-even ROAS Guide

Short answer

Normal break-even ROAS can be misleading for COD sellers.

Why?

Because COD revenue is not real at order placement. A placed COD order still has to pass confirmation, shipping, delivery, payment collection, and courier remittance.

A COD-adjusted break-even ROAS should calculate how much ad spend is safe after adjusting for:

• unconfirmed orders
• cancelled orders
• RTO orders
• courier charges
• COD collection fees
• packaging
• product cost
• discounts
• delayed cash collection

The practical rule:

Do not scale ads from dashboard ROAS until you know the ROAS needed on collected COD revenue.

Why this matters

A seller may see strong ROAS in an ad platform because the platform records a purchase when the COD order is placed.

But the seller may not collect that full revenue.

Some customers do not confirm. Some parcels are never delivered. Some orders return. Some cash arrives later. Some courier charges still apply even when the order fails.

This creates a gap between:

ad platform order value

and

real collected COD revenue

That gap can make a campaign look profitable while the business is losing cash.

COD/RTO funnel explanation

For prepaid ecommerce, an order often means payment has already happened.

For COD, an order is only the start of the funnel.

COD funnel stages

placed COD orders
→ confirmed orders
→ shipped orders
→ delivered paid orders
→ courier remittance received

RTO can happen after shipping:

shipped orders
→ failed delivery / refusal / unreachable customer
→ return to origin
→ cost already created

This means a COD seller should not calculate break-even ROAS from placed order value alone.

Formula logic

There are two useful ROAS views.

1. Dashboard ROAS

This is usually based on placed order value.

dashboard_roas = placed_order_value / ad_spend

This can be useful for measuring demand, but it can be dangerous for profit decisions if placed orders are not paid.

2. Collected COD ROAS

This is based on delivered and paid COD revenue.

collected_cod_roas = collected_cod_revenue / ad_spend

For COD sellers, this is usually the safer profit view.

COD-adjusted break-even ROAS

A simple break-even ROAS formula is:

break_even_roas = revenue / maximum_safe_ad_spend

For COD, revenue should be collected revenue, not just placed order value.

cod_adjusted_break_even_roas = collected_cod_revenue / maximum_safe_ad_spend

Where:

maximum_safe_ad_spend = collected_cod_revenue
                         - product_cost
                         - courier_delivery_costs
                         - rto_costs
                         - cod_collection_fees
                         - packaging_costs
                         - discounts
                         - confirmation_costs
                         - other_variable_costs
                         - target_profit

If target profit is zero, this gives a strict break-even point.

If target profit is included, this gives a safer scaling point.

Per-order version

If the seller wants a per-placed-order calculation, use expected values.

expected_collected_revenue_per_placed_order
= average_order_value × confirmation_rate × delivery_success_rate

Then estimate expected non-ad variable cost per placed order.

expected_non_ad_cost_per_placed_order
= product_cost_expected_on_delivered_orders
+ courier_cost_expected_per_placed_order
+ rto_cost_expected_per_placed_order
+ cod_fee_expected_per_placed_order
+ packaging_expected_per_placed_order
+ confirmation_cost_per_placed_order
+ discount_expected_per_placed_order
+ other_variable_costs

Then:

maximum_safe_cpa_per_placed_order
= expected_collected_revenue_per_placed_order
- expected_non_ad_cost_per_placed_order
- target_profit_per_placed_order

And:

cod_adjusted_break_even_roas_on_placed_value
= average_placed_order_value / maximum_safe_cpa_per_placed_order

This is useful when the ad dashboard counts placed COD order value.

Important warning

If maximum safe ad spend is zero or negative, the offer is not safe to scale at any ROAS.

That usually means one of these is true:

• product margin is too low
• RTO cost is too high
• delivery success is too weak
• discounts are too aggressive
• courier charges are eating margin
• CPA is already too high
• AOV is too low for the cost structure

In that case, the seller should improve the economics before adding more budget.

What data the seller needs

Order funnel data

• Placed COD orders
• Confirmed COD orders
• Shipped orders
• Delivered paid orders
• RTO orders
• Pending orders

Revenue data

• Placed order value
• Delivered paid order value
• Average order value by product or campaign
• Discounts
• Bundle or upsell value

Cost data

• Product cost
• Packaging
• Delivery charges
• RTO charges
• COD collection charges
• Confirmation or call center cost
• Damage or resale loss
• Transaction, bank, or remittance charges if relevant

Ad data

• Total ad spend
• Cost per placed COD order
• Campaign-level order value
• Campaign-level delivered paid value if available

Common mistakes

Mistake 1: Using normal ROAS for COD scaling

Normal ROAS may be based on placed order value. COD profit depends on collected revenue.

Mistake 2: Comparing COD ROAS with prepaid ROAS directly

A prepaid order and a COD placed order are not the same risk.

COD has confirmation risk, delivery risk, RTO risk, and cashflow delay.

Mistake 3: Ignoring RTO cost in break-even ROAS

RTO is not only lost revenue. It can create direct costs.

Mistake 4: Using countrywide assumptions

Do not use a fixed confirmation rate, delivery success rate, or RTO rate for all Pakistan, India, Bangladesh, UAE, or Saudi sellers.

Use your own data.

Mistake 5: Scaling because ROAS looks good

A high dashboard ROAS can still be unsafe if delivered paid revenue is low.

How SellMira helps

SellMira should help COD sellers move from surface-level ROAS to real profit logic.

A strong COD guide system should connect:

• placed order value
• delivered paid revenue
• RTO loss
• break-even CPA
• COD-adjusted break-even ROAS
• courier cost
• remittance delay

That supports SellMira’s positioning as a COD-first, ROAS-compatible profit leak engine.

Instead of telling sellers “your ROAS is good,” SellMira should help them ask:

Is this ROAS still profitable after COD risk?

Check your COD/RTO profit before scaling ads

Add forward shipping, reverse/RTO cost, COD fee, packaging, and ad spend to see whether placed COD orders are turning into collected cash.

Open COD/RTO Calculator

See a sample COD Profit Leak Report

Review how SellMira separates placed-order ROAS, collected COD ROAS, RTO loss, break-even CPA, and the first fix before scaling.

View Sample COD Report

Need a human check before increasing spend?

Request a one-time operational review of your COD funnel, RTO loss, courier fee logic, break-even CPA, and COD-adjusted ROAS.

Request Human COD Profit Audit

FAQ

What is COD-adjusted break-even ROAS?

It is the ROAS needed after adjusting for COD-specific risks like failed confirmation, RTO, courier costs, COD fees, and collected revenue.

Should I calculate ROAS on placed revenue or collected revenue?

For profit decisions, collected revenue is safer. Placed revenue can be used for demand tracking, but it should not be treated as cash.

Can my dashboard ROAS be profitable but COD ROAS be unprofitable?

Yes. This can happen when many placed COD orders do not convert into delivered paid orders.

Is there one good break-even ROAS for all COD sellers?

No. Break-even ROAS depends on your product cost, AOV, courier cost, RTO, delivery success, discounts, and target profit.

What should I do if my COD-adjusted break-even ROAS is too high?

Improve the offer economics before scaling. You may need better confirmation, better audience quality, higher AOV, lower courier cost, fewer discounts, or a product with stronger margin.

Source notes and caveats

• This guide does not claim default COD confirmation rates, delivery success rates, or RTO rates.
• Use seller-owned data and courier invoices.
• If a seller does not have final order statuses yet, pending orders should be excluded or tracked separately.
• Any public benchmark should be used only when verified and clearly cited.