Fashion Facebook Ads: Break-Even ROAS and What to Test First

Break-even ROAS for fashion Facebook ads is 1 ÷ (gross margin × (1 − return rate)). At a 55% gross margin and 25% of orders returned, you need 2.42x. At a 50% margin and 30% returns you need 2.86x. These are floors before shipping, payment fees and return handling, which all push the number up.

Gross margin is the share of the selling price left after product cost. Return rate is the share of orders refunded. This is our model, not an official figure. It assumes returned items go back on the shelf, refunds are full, and the platform reports ROAS on orders before returns.

The NRF figure covers all online retail, not only clothing. The rate that matters is your own, pulled for your hero product, the one item you put most of your ad spend behind. Sized garments and one-size accessories can return at very different rates, so calculate each product separately.

A worked example, with our assumptions: an $80 order at a 50% margin, 30% returns and a reported ROAS of 2.9x. You spent $80 ÷ 2.9 = $27.59 on ads for that order. Kept revenue is $80 × 0.70 = $56, and gross profit is $56 × 0.50 = $28. That leaves $0.41 per order before shipping, fees and returns handling. With no returns, the same order leaves $12.41.

A reported ROAS looks healthy because it records revenue at purchase. A later refund does not reduce it unless your store sends refunds back to the platform, while the ad cost stays paid. The difference between reported and kept revenue is your return rate.

So the question "am I actually making money at 1.5x ROAS?" has a precise answer. At 1.5x you would need a gross margin above about 67% with zero returns to break even on the first order. Almost no apparel brand has that.

Repeat purchases can change the picture, because a customer who buys again lowers the real cost of the first order. Track repeat rate separately. It is a reason to be patient with first-order results. It is not a reason to skip the break-even check.

If your reported ROAS sits below your break-even, look at the offer, the size guidance and the product page before changing creative. New ads will not repair a fit or returns problem.

Three 2026 apparel benchmark pages publish click-through rates from 1.95% to 2.84% and cost per acquisition from $10.95 to $31. They disagree because each uses its own sample and its own definitions, and some figures appear to be shared between pages, which is our inference. Treat the range as a warning, not a target.

Click-through rate (CTR) is clicks divided by impressions. Cost per acquisition (CPA) is spend divided by purchases.

Vendor-published figures, not independently verified:

Our arithmetic on those published figures: $31 ÷ $10.95 is about 2.8x between the highest and lowest CPA, and 2.84% ÷ 1.95% is about 1.5x for CTR. All three pages give the same $0.45 cost per click. Exact matches across independently measured samples are unlikely, which is why we read the shared figures as likely copied.

Set the ROAS column against the break-even table above. A benchmark ROAS of 2.9x to 3.2x clears a 2.42x break-even, but it only just clears 2.86x. Your own last 30 days, set against your own break-even, are the only benchmark that includes your margin. To check your CTR against your own history, see what a good CTR for Facebook ads looks like.

Read competitors' fashion ads as a record of what brands chose to run, not of what worked. A live ad shows that someone is paying for it. It does not show their cost per purchase or their return rate. Use it to build a list of ideas to test, and let your own account decide.

Our public Fall Fashion feed, built from the Meta Ad Library (Meta's public database of active ads), ranks creatives by how many variants each brand is running. A brand running many variants of one creative is investing in it, which makes it a better lead than a one-off ad.

For ten ads in your category, record five fields and apply these tests:

1. Format. Record whether the ad is a carousel, a video or a single image. Pass: you have all ten recorded, so you can see which format you have not tested.
2. First line. Copy the first line of copy or the first three seconds of video. Pass: you can sort the ten into no more than three angles, such as fit, occasion or fabric.
3. Offer. Mark whether the ad mentions a discount, code or free shipping. Pass: you know how many of the ten lead with price before you decide to.
4. Run time. Note the start date. Pass: you treat an idea as worth testing only if it appears in at least three of the ten and has run for four weeks or more. This threshold is our recommendation, not a measured rule.
5. Variants. Note how many versions of the creative are running. Pass: you can name the angle that gets the most variants.

Test format first, angle second, offer third and proof style fourth, one variable at a time, on your hero product only. This order is our recommendation. It puts the cheapest decision first and keeps the discount test until you know what the ad can say without one.

Our hypothesis for test 2, which the test is meant to check: aspiration ads draw more clicks, and fit-and-trust ads draw fewer returns. Judge it on return rate by ad where your store can report it, not on clicks.

Keep the audience, budget and landing page identical within each test. Write the hypothesis down before launch and decide in advance what result stops the test. A creative brief built around a single test keeps each round honest.

Start with two or three variants per test. Each variant needs enough purchases to read, so running twenty at once on a small budget leaves every one of them with too little data. That is our recommendation, not a measured threshold.

A fashion ad can describe the garment, the fit and the occasion. It cannot imply that the advertiser knows something about the person seeing it, and it cannot attack the viewer's appearance. Meta's personal attributes policy bans ads that "imply that the advertiser is aware of someone's personal attributes or the personal attributes of their family."

The policy lists race, ethnicity, religion, beliefs, age, sexual orientation or practices, gender identity, disability, physical or mental health (including medical conditions), vulnerable financial status, voting status, membership in a trade union, criminal record, and name. Meta's own example for apparel-type products: "We print customizable t-shirts and stickers with your name." is allowed, and "Billy Taylor, get this t-shirt with your name in print!" is prohibited.

Body and appearance claims fall under a second policy. As of 22 July 2026, Meta's Health and Wellness policy requires ads for dietary, health, weight-loss and weight-gain products and services to target people at least 18 years old. It permits before-and-after transformation images for general cosmetic products, procedures and surgeries when targeted to adults 18 and over. It bans "statements of inferiority about physical appearance," meaning terms, descriptions or questions that are negative and attack an individual's appearance, specific body parts or hygiene. That matters for shapewear, activewear and any fit claim.

Three checks to run on every line of copy before launch, as our recommendation:

1. Garment or wearer? Pass: the line describes the product ("high-rise, stretch denim in a relaxed fit"). Fail: it describes the reader ("for your problem areas").
2. Question about the viewer? Pass: no line asks the viewer about their age, body or health. Fail: any line does.
3. Body or appearance claims. Pass: no line attacks or implies a flaw in the viewer's body, and any weight-loss or cosmetic before-and-after content is targeted to 18+ and meets the Health and Wellness policy. Fail: copy such as "hide your problem areas," or any body-change claim you have not checked against that policy.

We build a ranked weekly test plan from four inputs: your ad performance data, competitor activity, customer signals and your past ad learnings. The plan lists which test to run next and why, and we turn the top ones into execution-ready creative briefs. Your break-even ROAS sets the bar each test has to clear.

For a fashion team, the four-test order above is a starting point, not a fixed script. Your account's own results decide whether format, angle, offer or proof style is the next gap. The competitor side uses public ad data like the feed above. The customer side reads what your buyers say, such as comments on size and fit, so the angle test starts from your customers' words.

The break-even math stays with you, because it depends on your margin and your returns. We rank what to test next, and you decide what result makes a test worth scaling. To size how many tests your budget can support, use our creative test budget calculator. Plans and prices are on our pricing page.

If you want to see this on your own ads, book a demo with our founding team.