Returns and ad profitability: the e-commerce blind spot

Returns reduce ad profitability because platforms count the sale at checkout and never subtract the refund. A campaign with a high return rate can look like the best performer on platform ROAS while delivering the least profit. Tying returns back to the campaign and creative that drove each order shows which ads attract buyers who keep what they buy.

Why are returns the biggest blind spot?

Returns arrive days or weeks after the purchase event, long after the ad platform has credited the conversion and the bidding algorithm has learned from it. The algorithm is effectively trained to find more people like the ones who bought, including those who sent things back. If a creative attracts impulse buyers with high return rates, the platform rewards it and spends more on it.

For many fashion, footwear and home brands, returns are the largest single gap between platform ROAS and true profit, larger than fees or shipping. We cover the full calculation in our true-profit ROAS guide; this article is about the returns line specifically.

How do you tie returns back to ads?

1. Keep the source on the order. UTMs, click IDs or first-party tracking stored with each order.
2. Join returns to orders. Most store and returns platforms can export by order ID.
3. Aggregate by campaign, then by creative. Return value as a share of attributed revenue.
4. Split by product category. Some campaigns look bad only because they sell high-return products.
5. Refresh monthly. Returns lag; recent weeks will be understated until the return window closes.

Where matching is incomplete, apply category return rates to each campaign's product mix. It is an estimate, but a labelled estimate beats an invisible cost.

What patterns usually show up?

• Creatives that oversell: flattering angles, ambiguous sizing, heavy filters. More clicks, more returns.
• Discount-heavy campaigns: lower commitment buyers who return more readily.
• Broad prospecting at saturation: marginal buyers with weaker intent, as described in our diminishing returns guide.
• Specific products: fit-sensitive items that return heavily regardless of campaign.

What can marketing change?

More than you might expect. Returns are often treated as an operations problem, but ads set expectations. Practical levers include:

• Showing the product honestly: real fabric, real fit, people of different builds.
• Putting sizing guidance in the creative or first landing view.
• Reducing discount depth on campaigns whose buyers return most.
• Moving budget towards creatives with lower return-adjusted cost per kept order.
• Excluding repeat serial returners from retargeting audiences where policy allows.

This is where honest creative and profitable creative turn out to be the same thing. An ad that earns a click by setting accurate expectations tends to earn a customer who keeps the product.

Definitions

• Return rate: returned value as a share of sold value in a period.
• Return-adjusted CPA: ad spend divided by orders that were kept.
• Return window lag: the delay before returns for a period are fully known.

How does audience insight help reduce returns?

Return reasons are a rich source of audience truth: "smaller than expected", "colour different", "not as pictured". Read alongside what people say about a category in social conversation, they tell you which expectations your creative must set. SOMIN, an AI audience-research platform, decodes the tensions and tones in what real audiences post; SOMIN for brands outlines how brand teams use that. The Miraflora case study on somin.ai is a good example of brand work grounded in audience evidence.

For the content side, our sister consultancy Mindfuse works on the consumer psychology behind purchase decisions in Southeast Asia, including the cultural expectations that shape what buyers assume a product will be.

A worked example: two creatives, one product

Consider two illustrative creatives selling the same dress. Creative one is a glossy studio shot with a dramatic angle; it earns a high CTR and the most purchases. Creative two shows the dress on two different body types in daylight with a visible size guide; it earns fewer clicks and somewhat fewer purchases. On platform ROAS, creative one wins clearly. Then returns arrive. Creative one's buyers return a large share, often citing fit and colour. Creative two's buyers keep most of what they ordered. Once returns, return shipping and restocking costs are netted out, creative two delivers more profit per unit of spend, and its return-adjusted CPA is lower. The algorithm, left alone, would have kept funding creative one.

The lesson is not that glossy creative is bad. It is that clickability and keepability are different things, and you need to measure both. Score for clickability before launch; judge on kept orders after it.

What about returns on Google Shopping?

Shopping and Performance Max campaigns are driven by product feeds, so returns there often concentrate on specific products rather than creatives. Review return rates by product ID, and consider lowering bids or excluding persistent high-return items from paid campaigns until the product page or sizing information is fixed.

The cheapest return to process is the one your ad never caused.

How FMedia tracks returns

FMedia joins returns data to orders and orders to campaigns and creatives, then shows return-adjusted figures next to platform ones: return rate, return-adjusted CPA and true-profit ROAS. Recent weeks are marked as provisional until the return window closes, so you do not overreact to incomplete data. Creatives with high return rates are flagged, and their replacements can be scored before launch for clickability, with the brief adjusted to set clearer expectations.

The goal is not fewer sales. It is more kept sales per unit of spend, which is the only kind that pays the bills.

Frequently asked questions

Do Meta and Google subtract returns from ROAS?

No. Both platforms count the purchase value at checkout. Returns and refunds must be subtracted using your own order and returns data.

How long should I wait before judging a campaign's returns?

Until your return window has closed for that period, often a few weeks. Treat earlier figures as provisional.

Can ad creative really reduce return rates?

Often, yes. Creatives that show products honestly and set clear expectations on fit, colour and finish tend to attract buyers who keep what they buy.