Facebook ads for clothing brands

Start with the reach, because it is the whole reason to be here. Pew's November 2025 numbers put Facebook thirty-four points ahead of Instagram among 50- to 64-year-olds (74% against 40%) and thirty-eight points ahead among over-65s (57% against 19%). No other industry in this cluster gets that much from the platform axis, because for most trades an older customer is simply an older version of the same customer. In apparel they are a different customer with different problems: a body that has changed shape, a size range the high street has narrowed, a favourite garment that has been discontinued, and thirty years of learning which brands lie about fit.

That is a strong position and it comes with a specific tax. Meta's Advertising Standards prohibit ads that assert or imply personal attributes about the person seeing them, and the list includes age, disability, and physical or mental health. Every sentence that names why this customer needs you lands on it. "Trousers for women over 50" asserts age. "Since your hip replacement" asserts a medical condition. "Struggling with wide feet?" asserts a physical characteristic and does not stop being an assertion because it is phrased as a question — the question form is not an exemption anywhere in this policy, which is exactly the trap the dental vernacular falls into on our dentist pages. Meta does permit passing reference to an attribute, including age ranges, but the line is whether the ad is telling the reader something about themselves, and retail copy written to this audience almost always is.

Which produces the inversion with our Instagram guide, and it runs in the direction most people would guess wrong. That page says to go broad in the ad set and let the creative narrow the audience, because no picker has ever offered "women who are five foot ten". Here the narrowing you want is age, apparel is not a special ad category, and age is fully available in the picker — so the qualifying moves out of the creative and back into the ad set. Facebook is the app where you may select the customer and may not describe them. That single sentence organises the rest of this page.

The mechanical differences are smaller but they compound. Meta's ads guide recommends a 27-character headline for a Facebook Feed image ad against Instagram Feed's 40, and 50–150 characters of primary text against 125; both re-read from the spec pages on 7 September 2026. Facebook Feed shows no description line at all while Marketplace shows 30 characters. And Advantage+ placements will fill nine image surfaces from one export, so the 4:5 you designed for the feed will appear as a square on Marketplace and small in the right-hand column whether you planned for it or not.

The 27-character limit is worth a paragraph of its own, because our own batch measured it and the result is the sharpest in this cluster. Ten ads produced fifty headlines. Two of them fit 27. Forty-eight fit Instagram's 40. All ten lead headlines overran Facebook's recommendation, which is worse than the four Facebook batches before it — those ran eight or nine of ten over. Every specification headline overran too: "E to 6E. Cut once. Still available." is 35 characters and "0.4% shrinkage after 40 washes." is 31. The two that fit were "The tag names the frame" and "Knit where we spin the yarn". Twenty-seven characters is not room for a claim. It is room for a noun.

Sales, same as the Instagram twin, and the fact that the answer does not move is itself the finding. Our Facebook batch is five Sales, three Traffic, one Leads and one Awareness, against Instagram's six, three, one and none. The SaaS cluster is the only one on this site whose objective mix flips across the platform axis, and it flips because the label "SaaS" covers two business models with two different buyers. Apparel is one business model. The buyer changes; the mechanic does not. A purchase still completes in the ad's own session, so Sales still describes the thing that can actually happen.

What does change is how much convincing happens before the checkout. This reader has been burned — by fit, by sizing charts, by a parcel that took six weeks from an address nobody would print — so a larger share of the work is arguing rather than selling, and Traffic is the honest objective for the ads doing it. The three Traffic ads on our Browse page are all of that kind: one explains that vanity sizing drifts and the labels are the variable, one publishes a wash test, and one makes the case that a return address is a diagnostic you can apply to every other ad in the feed.

Leads is more real here than on Instagram, and for a reason worth knowing: this buyer will still give you a postal address. A swatch card, a fabric sample, a printed lookbook — these are lead mechanics that the Instagram audience would find quaint and that work on this app, because they suit a considered purchase and a reader who is not going to choose a £395 cloth from a phone screen. That is the one Leads ad on our examples page, and it is the only offer in this cluster where the deliverable arrives in the post.

Apparel has no service menu, so the unit of planning is an argument rather than an offer — and on this app every one of them has to be an argument about a garment. That is not a stylistic preference, it is the policy rule from the platform section applied to the planning table: the moment a row is about the reader's body, age or health rather than about what you make, it stops being an ad you can run. Read the "who" column below as a targeting note and the row itself as the thing the ad is allowed to say.

Each row pins four decisions: who it is for, the angle family that argues it (each is a page in the angle library), the objective to buy it under, and the finished ad on the Browse page that shows it assembled. Ten of the twelve have an example, because the batch ran all ten angle families once each — including the Founder POV row the Instagram twin left empty. Two rows have no example on purpose and both are named below.

One recurring shape worth naming before the table: nearly every row is about getting the same thing again. Reorder, restock, resole, re-cut, the block that has not moved since 1998, the last cut in 2004. On Instagram the strongest ads narrow by body — tall, petite, big and tall — because that reader is discovering. Here the strongest ads narrow by continuity, which is a property of your product rather than of your customer, and that is the same distinction the policy is drawing.

Apparel is not one of Meta's special ad categories, so every lever survives — and on this app that matters more than it does on the twin, because the lever you most want is the one Instagram cannot deliver an audience for. Age is selectable. The 50–64 and 65+ bands are where this buyer is, and reaching them is an ad set decision rather than a creative one. Note the asymmetry that produces: you may target by age and you may not write about it, so the ad set and the copy are doing two different jobs, and only one of them is allowed to know who it is talking to.

Since 31 March 2020 every campaign created through the Marketing API has had to declare a special_ad_category even when the answer is NONE, so declare it and move on. The one thing to watch is that the category attaches to what an ad offers rather than to what a company sells: a creative promoting instalments, a store card or financing is a financial-products ad regardless of the fact that you sell jumpers, and that category carries housing's restrictions — fixed 18–65+ age, no lookalikes, no saved audiences, a forced radius. If you run a finance offer, that ad set loses the age lever this whole page is built on.

Everything else is ordinary and the ordinary answer is right. Go broad within the age band and let the garment do the rest — the segmentation that matters here is by body and by habit, and no picker has ever offered "wears a 4E" or "has three sizes of the same trouser". Exclusions are available to you, unlike in housing, and the two worth setting are recent purchasers on acquisition ad sets and serial returners on everything.

Two things this page will not tell you, because both are the sort of claim that gets repeated across marketing blogs until nobody remembers where it came from. First, which specific detailed-targeting interests currently exist for clothing or shopping behaviour: Meta has narrowed detailed targeting repeatedly and the honest instruction is to open the picker in your own account. Second, whether a given interest "still works" — the recommendation is to go broad regardless, so the answer changes nothing.

The last lever is the one everybody forgets on this app: placement. Advantage+ will put your 4:5 feed export into Marketplace as a square and into the right-hand column at a size where a paragraph of body copy is a smear. Either accept that and design a composition that survives it, or turn placements off for the ads where the frame carries the argument — which for apparel is most of the good ones.

The product page for the item in the ad, same as the Instagram twin, and for the same reason: an ad that argued about one inseam landing on a grid of forty trousers has handed the reader back the search problem it just solved. What changes here is what has to be on that page. On Instagram the two things to put above the fold are the size guide and the returns policy. Here it is those two plus the things that prove you exist — a street address, a phone number, a returns address in the same country, a company registration. This reader is checking.

Do not put the size guide behind a modal or ship it as an image. This is the buyer most likely to be reading on a large screen with text scaled up, and a size chart rendered as a JPEG is unreadable at 150% and invisible to a screen reader. It is also the single most-visited page on an apparel site for this audience, and treating it as a footnote is the most common own goal in the category.

Lead forms are worth more here than on the twin and only for something that physically arrives: a swatch card, a fabric sample, a printed catalogue. Use an Instant Form on Higher intent so there is a review step, ask for a postal address because that is the whole point, and do not run it if you are not actually going to post anything. A lead form standing between this reader and a checkout is a step that costs a sale and buys an email address.

One placement-specific note. If Advantage+ is on, some of your traffic will arrive from Marketplace, which is a shopping surface with its own expectations — people there are browsing listings, not reading brand pages. A click from Marketplace landing on a lookbook converts worse than the same click landing on a priced product page with a delivery date on it.

The first rule is the policy rule and it is a creative rule, not a legal footnote: the ad talks about the garment and never about the reader. Every argument you want to make has a version pointing at the product and a version pointing at the person, they select exactly the same audience, and only one of them can run. "E to 6E, resoleable" instead of "for wide feet". "Finished waist in inches" instead of "if your size keeps changing". "Machine wash cold, tumble dry low" instead of "nobody has time for the dry cleaner". Write the product version first and you will rarely have to catch the other one.

Our own batch is the evidence that this works at the brief rather than at the review. Ten ads were written to that rule and not one came back naming an age, a body or a condition — which is worth contrasting with the gym batch on this site, where ten briefs written to avoid body transformation still produced "SAME BODY" rendered into an image, because there the offer itself was about the body. If the thing you are selling is a garment specification, the copy has nowhere else to go. If it is a transformation, no amount of steering keeps it away.

The second rule is the one the same batch failed, five times out of ten, and it is the expensive one: where a direction offers a place to put proof, the model fills it. A wash-test logbook came back with four rows of shrinkage percentages nobody measured. Two ads spoke in the first person as customers who do not exist. A gift ad rendered a private text conversation between two named siblings about their mother. A linen ad — whose entire argument is that its return address is real — stamped a specific Los Angeles garment-district address on the parcel. None of that was asked for and all of it is the same failure: an evidence slot in the composition and no evidence to put in it. If you cannot substantiate a number or a person this week, choose a direction with nowhere to put one.

Then the frame, which on this app is not the frame you designed. Advantage+ placements fill nine image surfaces from one export, so compose 4:5 for the feed with a square crop out of the middle that still works. Check the re-framed file rather than the square draft: two of the nine non-square exports in our batch arrived with the original picture inset inside the new canvas and white bars top and bottom, and on 1.91:1 — which routes to the smallest surfaces on the app — that costs a sixth of an already small asset.

Three stages, and the differences from the Instagram twin are all at the ends. At the top, more of the budget goes to establishing that you are real and that the sizing is honest, because this reader will not buy from a brand they have not verified — which is why our batch is Traffic-heavier at the front than a pure Sales structure would be. At the bottom, the customer stage is worth more than it is on the other app: a buyer who has found a garment that fits will reorder it for years, and that is the cheapest revenue in the account.

The middle stage is where the apparel-specific work sits, and it is objection handling rather than reminding. Somebody who looked at a product page and left almost always left for one of two reasons — they were not sure it would fit, or they were not sure about you. Retarget with the answer to each rather than with the same product photograph and a discount, which reads as a brand that was overcharging the first time.

The learning-phase arithmetic is the same everywhere and lands differently for every business: roughly 50 optimisation events per ad set per week to leave the learning phase, so multiply your own cost per purchase by 50 and that is the weekly spend at which one ad set is fully optimised. For apparel it turns entirely on average order value — 50 purchases a week is routine at €40 and a serious budget at €400. If the number is out of reach, consolidate ad sets rather than splitting the audience further, and accept that you are reading a fortnight rather than a week.

The specific temptation on this app is to keep adjusting the age band, and that is a significant edit: it restarts the learning phase in the same way a new optimisation event or a new creative does. Narrowing 45–65+ to 50–64 because the reports look better in one row costs you a fortnight of delivery to buy a distinction the creative was never allowed to make anyway.

The calendar itself is different on this app and it is the part most apparel media plans get wrong. Instagram's fashion calendar is drops and seasons. This one is gifting: Christmas, Mother's Day, Father's Day, and the long tail of birthdays that a gift-with-exchange-card offer serves all year. Auction prices in the six weeks to Christmas are the highest you will pay, and the honest move is the same one the gym guide makes about January — buy the month before the peak, when the same customer is deciding and nobody is bidding against you.

Finally the number Ads Manager cannot show you, and it is worse here than on the twin. Returns in apparel arrive weeks after the sale and never appear in reported ROAS, and this audience returns more than a younger one does, for the specific reason that fit is less predictable on a body the industry does not cut for. So reconcile on net revenue after returns on a three-to-four week lag before scaling anything, and read a high return rate on one ad as a creative fault: an ad that oversold the fit is what produced it.

The first two are the ones that get accounts restricted rather than ads rejected, and they are both consequences of writing the sentence that comes naturally.