True Ecommerce ROAS After Returns, Refunds and Cancellations

A campaign can report strong ROAS and still produce weak profit after refunds, returns and cancellations are included.

Advertising platforms usually optimize from the conversion value sent at purchase time. They may not fully reflect what happens weeks later when an order is refunded, partially returned or cancelled.

For ecommerce brands, the correct question is not only:

How much revenue did the ads generate?

It is:

How much net contribution remained after customer returns and advertising cost?

Why platform ROAS can overstate the result

Platform ROAS is commonly calculated as:

Attributed purchase revenue divided by ad spend

If a campaign spends $10,000 and reports $40,000 in purchase revenue:

$40,000 divided by $10,000 = 4.00x ROAS

If $8,000 of the orders are later refunded, retained revenue becomes $32,000.

$32,000 divided by $10,000 = 3.20x revenue-after-refunds ROAS

The result falls before product cost, shipping, processing and return handling are included.

Define every post-purchase loss

Separate:

1. Full refund
2. Partial refund
3. Product return
4. Exchange
5. Order cancellation
6. Chargeback
7. Failed delivery
8. Reshipment
9. Damaged item replacement
10. Goodwill credit
11. Store credit
12. Subscription cancellation

These outcomes have different financial effects.

A return may recover the product for resale. A damaged replacement may lose both product and shipping cost. Store credit may delay rather than eliminate revenue.

Calculate net retained revenue

A practical starting formula is:

Net retained revenue = paid order revenue minus refunds minus cancellations minus chargebacks

Document whether the figure includes:

1. Shipping revenue
2. Tax
3. Gift cards
4. Store credit
5. Partial refunds
6. Currency conversion

Use the same definition every reporting period.

Calculate expected return cost per order

Historical cohorts can estimate the cost before every return is fully mature.

Expected return cost per order = return rate multiplied by average loss per return

Example:

If return rate is 10 percent and average financial loss per return is $35:

0.10 multiplied by $35 = $3.50 expected return cost per order

Average loss per return may include:

1. Refunded revenue
2. Non-recoverable product cost
3. Outbound shipping
4. Return shipping
5. Payment fees not recovered
6. Warehouse handling
7. Inspection
8. Repackaging
9. Discount required for resale
10. Customer-service cost

Calculate contribution after returns and ads

Use:

Contribution after returns and ads = net retained revenue minus variable product costs minus return costs minus ad spend

Example:

The platform may still display 4.00x ROAS, but the business retained $6,000 before fixed operating costs.

Use mature cohorts

Returns often happen after the advertising reporting window.

Compare orders by purchase cohort.

For example:

1. Orders placed in January
2. Refunds and returns related to those January orders
3. Final retained revenue after the return window matures

Do not compare January ad spend with all refunds processed in January, because some refunds may relate to December orders.

A cohort table:

Review return rate by product

Campaign averages can hide problem products.

Track:

Common causes of high returns include:

1. Sizing problems
2. Inaccurate product images
3. Misleading creative
4. Weak quality
5. Compatibility confusion
6. Shipping damage
7. Product expectations not matched
8. Variant errors
9. Delayed delivery
10. Impulse-buy promotions

Review return rate by creative concept

Different ads can attract different customer quality.

A creative may increase purchases while increasing returns because it:

1. Overpromises the outcome
2. Hides product size
3. Shows an inaccurate color
4. Uses unrealistic demonstrations
5. Emphasizes a use case the product does not support
6. Creates urgency without qualification
7. Attracts gift buyers who misunderstand the product
8. Promotes the wrong variant

Track:

A higher initial CPA can be better when it creates customers with lower return rates and stronger retained contribution.

Review return rate by channel

Compare:

1. Meta Ads
2. Google Search
3. Google Shopping
4. Performance Max
5. Email
6. Organic Search
7. Direct
8. Influencer
9. Affiliate
10. Marketplace

Do not assume channel differences are caused only by targeting. Product mix, offer and customer intent may differ.

Review new and returning customers

New customers may return at a different rate from returning customers.

Track:

1. New customer return rate
2. Returning customer return rate
3. First-order refund rate
4. Repeat-order refund rate
5. New customer retained CPA
6. Contribution after returns
7. Payback period

A campaign with a low purchase CPA may have a high retained-customer CPA.

Calculate retained customer CPA

Retained customer CPA = acquisition spend divided by new customers who remain after full cancellation or refund

Example:

If the business spends $20,000 and acquires 500 new customers, reported new customer CPA is:

$20,000 divided by 500 = $40

If 50 customers fully refund or cancel:

$20,000 divided by 450 retained customers = $44.44 retained customer CPA

This still does not include partial-return quality, but it gives a more realistic acquisition view.

Review geography

Return economics can vary by market because of:

1. Shipping time
2. Return shipping cost
3. Duties
4. Product expectations
5. Sizing standards
6. Currency
7. Local return rights
8. Carrier damage
9. Failed delivery
10. Customer support coverage

A country with strong initial ROAS may lose profit after international returns.

Review promotion quality

Large discounts can increase both conversion and low-intent orders.

Compare:

1. Full-price orders
2. Small discount orders
3. Major promotion orders
4. Free-gift orders
5. Bundle orders
6. Subscription discounts

Track:

1. Conversion rate
2. CPA
3. Average order value
4. Return rate
5. Refund value
6. Contribution after ads
7. Repeat purchase

A promotion should be judged by retained profit, not only sales volume.

Improve product-page expectation setting

Reduce avoidable returns with:

1. Accurate images
2. Size and dimensions
3. Materials
4. Compatibility information
5. What is included
6. Usage instructions
7. Delivery estimates
8. Honest limitations
9. Reviews
10. Clear return policy

A higher-quality purchase is usually more valuable than an easier purchase that later reverses.

Improve creative accuracy

Before launching, confirm:

1. Packaging is current
2. Product scale is clear
3. Color is accurate
4. Variant shown is available
5. Claims are supported
6. Demonstration is realistic
7. Bundle contents are correct
8. Price and offer are correct
9. Shipping message is accurate
10. Customer outcome is not guaranteed without basis

Creative accuracy is a profitability control.

Feed return data into media decisions

Use return-adjusted labels where reporting systems allow.

At minimum, classify products and campaigns by:

1. Low return rate
2. Average return rate
3. High return rate
4. Return reason
5. Contribution after returns
6. Customer quality

Actions may include:

1. Reduce spend
2. Fix the product page
3. Change creative
4. Exclude a market
5. Change sizing information
6. Improve packaging
7. Adjust the CPA target
8. Pause the product

Build a return-adjusted scorecard

Use the same table by product, creative and market.

Set operating targets with expected returns

Do not wait for returns to surprise the business.

If a product historically loses 8 percent of revenue to refunds and cancellations, include that expected loss in the contribution model before setting maximum CPA.

Review the assumption regularly because return rate changes with:

1. Product updates
2. New creative
3. Promotions
4. Market expansion
5. Shipping changes
6. Seasonality
7. Sizing changes
8. Customer mix

Common mistakes

Comparing refund processing date with ad spend date

Different order cohorts are mixed.

Looking only at full returns

Partial refunds and reshipments are ignored.

Applying one return rate to every product

Problem products remain hidden.

Blaming the channel without checking product mix

One channel may sell more of a high-return product.

Optimizing only to purchase events

Customer quality after purchase is ignored.

Using gross revenue in the break-even model

Maximum CPA is overstated.

Final principle

True paid ads performance is measured after the customer keeps the order.

Reconcile refunds to the original purchase cohort, calculate retained revenue, include return handling and product loss, and compare contribution by campaign, creative, product and market. Scale the acquisition that creates retained customers and retained profit.

Continue with the ROAS versus net profit guide and product profitability segmentation guide.

Book a free Beelog profit audit to calculate return-adjusted ROAS, retained customer CPA and product-level contribution.