Instagram ads for gyms and fitness studios

Fitness content on Instagram is transformation content, and it is posted for free by people who are not advertisers. That is the platform fact that matters more than any spec. The Health and Wellness standards prohibit ‘statements of inferiority about physical appearance (i.e., terms, descriptions, or questions that are negative and attack an individual's appearance, specific body parts, or hygiene)’, the ‘close up on specific body area by pinching fat’, and — in a health, weight-loss or weight-gain context — ‘sensational language with exaggerated or extreme claims, or promises of specific outcomes within a set timeframe without disclaimers or qualifiers’. Line those up against the genre and most of it fails: the before shot works by being negative about a body, the caption promises the after by a date. A creator can post it. You cannot pay to.

Which turns out to be the argument rather than the obstacle. What a gym has that a creator does not is a building and a contract: a rack that is free at six, a door that opens at two in the morning, a class pack that does not expire, a coach whose rehab background is printed on the wall, a membership that cancels in two taps. None of the ten ads on the Browse page sells a body, and every one of them sells something the reader cannot get from the transformation post above it. In a feed where the results claim has been discounted to zero by repetition, the schedule is the differentiated message.

The second thing Instagram decides is which of your two businesses it favours. Pew's November 2025 report puts Instagram at 80% of adults aged 18 to 29, 62% of 30- to 49-year-olds and 40% of 50- to 64-year-olds, against Facebook's 68%, 80% and 74%. A boutique studio's core member sits squarely in the first band; an access gym's membership spreads across all three and its price-led, cancel-anytime half skews older than the studio's. So this app leans studio, the Facebook twin of this page will lean access gym, and a gym that is genuinely both should expect the two platforms to fill different halves of the floor.

Third, and it bears repeating because two of the four guides before this one open with the opposite: none of this is a special ad category. Meta's Marketing API accepts exactly four — housing, employment, financial products and services, and social issues — so a gym keeps every lever a real estate agent loses, and the custom-location radius in the API reference goes down to 0.63 miles, or one kilometre. A gym's addressable market is the people who will physically walk in, and unlike the housing pages in this cluster you can actually buy that shape.

Finally the shapes and the copy lengths. Meta recommends 4:5 at 1440 × 1800 for Instagram Feed, identical to Facebook Feed so one export serves both, with an aspect-ratio tolerance of 1% here against 3% there — a crop a few pixels out passes on Facebook and fails on Instagram from the same file. Eight of the ten Browse ads are 4:5 and two are 9:16 for Stories and Reels, and none is 1.91:1 — a choice rather than a rule, corrected here on 7 September 2026: Instagram Feed's technical requirements accept aspect ratios from a minimum of 4:5 up to a maximum of 1.91:1, so what this app lacks is not the shape but the surfaces built to want it, which are Facebook's right-hand column and search results. The headline recommendation is 40 characters and the primary text 125. Measured against that, one of the ten headlines on the Browse page is over and all ten primary texts are, at 276 to 445 characters — so the argument has to land inside the first sentence, and on this page the first sentence is usually a time of day.

Ads Manager offers six objectives, and a gym's answer depends on which of the two businesses it is and on one question that sounds administrative and is not: does the thing you are selling complete in a browser? A class pack does — it is a fixed-price product bought at a checkout. A membership that needs a tour, a signature or a conversation does not. An access gym whose join flow finishes in the app sits between the two and gets to choose, which is a luxury the dentist and the real estate agent in this cluster do not have.

The Browse page's ten were pinned as 5 Leads, 3 Traffic, 1 Sales and 1 Awareness, which places this trade exactly where that question predicts: between the local-service clusters, which are almost entirely Leads, and the apparel cluster, which is Sales-led. Leads is the default because most of what a gym sells is a first visit — a trial class, a re-start session, a tour — and the ad's job is to get one booked. Traffic is for the arguments that need a page to be believed, and there are more of those here than in most trades: how an unstaffed door works, what the occupancy counter shows, why two strength sessions beat five on a treadmill.

Sales is the interesting one. It is buyable for the class pack and the intro offer, and for an access gym whose join completes online — a real purchase event firing at a checkout is the best signal this platform can optimise on, and a studio with a $260 pack is the only local business in this cluster that has one. Below the volume the learning phase needs it is a worse Traffic campaign; the budget section has the arithmetic for which side of that line you are on.

This is the section the merged page has to get right, because the two businesses under this URL have almost no offers in common. A boutique studio sells a coach, a capped room and a schedule, and its unit of sale is a class or a pack of them. A 24/7 access gym sells a building and a contract — price, hours, how easy it is to leave — and its unit is a rolling membership. The first five rows below are the studio's, the next four are the access gym's, and the last three fit either. The ‘who it is for’ line on each row says which, and the angle recommendation changes between them: a studio's best arguments are proof and story, because it is selling a person; an access gym's are objection-handling and mechanism, because it is selling a building the reader has doubts about.

Each row pins four decisions: who buys it, the angle family that argues it (each a page in the angle library), the objective to buy it under, and the finished ad on the Browse page that shows it working. Ten rows have an example and two do not; the two are the intro offer and the off-peak membership, which the batch skipped because they need a real published price to be anything other than a placeholder, and which are the rows you will write yourself on the first day.

One rule runs across all twelve. Not one of them sells a result — not because results are a weak argument, but because the rows that never mention weight keep the ad set out of the 18+ floor and the ad out of the standards' clickbait clause, and because every competitor in the feed is already making the results claim for free.

Start with what you keep, because it is everything. Meta's Marketing API accepts exactly four special ad categories — housing, employment, financial products and services, and social issues, elections and politics — and fitness is not among them. So age is yours, the radius is yours down to the API's custom-location floor of 0.63 miles or one kilometre, lookalikes and exclusions are yours. If a guide tells you to leave age open because ‘Meta restricts health advertisers’, it has confused you with a real estate agent, exactly as the dental guides do.

Now the rule that is specific to this trade, and it is conditional, which no other industry on this site is. The Health and Wellness standards say that ‘ads promoting or marketing dietary, health, or weight loss or weight gain products and services must be targeted to people at least 18 years or older’ — and then list what ‘the age targeting restrictions above don't apply to’, which includes ‘products and services marketed for general wellbeing, such as fitness services and equipments, health clubs’. A gym is on the exempt side of that line for as long as its ad is about training, schedule, coaching and access. Put ‘lose weight’ in the primary text and the same gym is on the other side: the ad set must be 18+, and the copy is inside the standards' clickbait clause about exaggerated claims and outcomes promised within a timeframe. Our own examples page first read that as a blanket floor; read in full, it is a floor you choose to stand under by what you write. That is the most practical sentence on this page — the creative sets the targeting rule, so decide which ad you are writing before you open the ad set.

The rule that applies on both sides of the line is the one that caught the Browse page's own output: no ‘statements of inferiority about physical appearance’ — terms, descriptions or questions that are negative about a person's appearance, specific body parts or hygiene. Meta's personal attributes policy adds that an ad may not assert or imply a person's physical or mental health, directly or by implication, so ‘Out of shape?’ fails twice. As with the dentist, the audience brief and the ad copy are two different documents. ‘People still paying for a gym they stopped attending’ is a fine targeting instrument and a fine first line, because it describes a subscription and not a body. ‘Still carrying the January weight?’ describes a body.

Two practical notes. A member list is not protected health information the way a patient list is, so it is a usable seed and a usable exclusion — but Meta's Business Tools terms still prohibit data based on health information, including the names given to events, conversions and custom audiences. ‘Weight Loss Enquiry’ as a custom conversion is the shape that gets flagged; ‘Trial Booked’ is the same event named for what happened. And geography for an access gym is a route rather than a circle: the night-shift reader trains between work and home, so a pin on the building and a pin on the industrial estate two miles away are both your catchment, and a single wide radius is the lazier version of both.

Three answers, and the business decides. For the studio, the honest one is the DM. A boutique studio already lives in Instagram messages — that is where its members ask whether a class is full and whether the 6:15 is beginner-friendly — and Meta's lead ads can send people straight into a conversation on Messenger, Instagram or WhatsApp, with an ‘Instant forms and Messenger’ conversion location that runs one campaign and lets delivery pick the form or the chat per person. The condition is staffing: a click-to-message ad with nobody on the thread for four hours is paid-for silence. If the front desk answers messages inside the hour during opening hours, buy the conversation; if it does not, buy the form.

The form, when you buy it, should be the Higher-intent type. Meta's two Instant Form types differ by one step: ‘More volume’ submits in a thumb movement from pre-filled profile data, ‘Higher intent’ adds a review screen. The review screen is a filter you are paying for deliberately — fewer leads, more of them who remember filling it in when the coach rings — and for a trial class, where the cost of a no-show is a spot somebody else wanted, that is the right trade. Add one question that matters: which class, or which time of day.

For the access gym it is usually a page, for two different reasons. If the join flow completes online, send the click to it and let Sales optimise against the purchase — the cancel-anytime ad and the price lock both want the reader to do the thing now. And the arguments — the unstaffed door, the occupancy counter — need a page to be believed before anyone will hand over a card; a form asks for the commitment before the doubt has been answered. Class packs and intro offers are a checkout, full stop.

Never the homepage. A reader who tapped an ad about four private sessions before the main floor and lands on a rotating hero of a spin class has been asked to start their search again, and the whole point of the ad was that they would not have to.

The rule that decides everything else is one the Browse page learned the hard way. Every one of its ten briefs was written to sell schedule, contract, coaching and access rather than a body, and the single brief that came closest to training outcomes came back with ‘TEN YEARS ON THE TREADMILL. SAME BODY.’ and ‘Neither does the mirror.’ rendered into the image, where no alternate headline can reach it. Steering the brief is not the same as controlling the output, and this is the category where the gap is expensive: that line is the statement of inferiority about appearance the standards name, and it is baked into the pixels. The fix is small and general — ‘same numbers for years’ is about the training, ‘same body’ is about the person. Make every line the first kind, and check the image, not just the copy.

The second rule follows from the targeting section: decide which ad you are writing before you write it. A fitness ad and a weight-loss ad are under different rulebooks and the word that moves you between them is small. The ten Browse ads never use it — not as a dodge, but because the argument a gym can make that a creator cannot is about the building and the timetable, and that argument does not need the word.

Third, the test the cleanest ad on the Browse page passes: would it survive with the photograph removed? A solid-colour card that says ‘The confusion happens in private’ is an ad for a gym with no body in it, and it would comply with rules that catch most of the category. If your ad only works because of what the body in it looks like, it is the genre you cannot run, in a nicer font.

Three stages, and the third is the one most gyms already own and never advertise to. Cold buys the first visit, warm answers the doubt, and the membership list is both an exclusion and — split into current and lapsed — two audiences that behave nothing alike. Lapsed members are the cheapest acquisition in the building: they know where the machines are, they left for a reason that may have changed, and the re-start ad was written for them before it was written for anyone else.

Keep it this simple on purpose. A single-site gym's budget is small against the evidence Meta's delivery system wants, and every extra ad set divides the same evidence into thinner piles. Variation belongs at the ad level — the studio ad and the access-gym ad live in the same ad set if you run both businesses — not at the ad-set level.

Do the learning-phase arithmetic before anything else, because it decides how much of Meta's optimisation you are actually buying. An ad set needs roughly 50 optimisation events a week to leave the learning phase; below that it sits in ‘learning limited’, which is not a penalty but is the system saying it cannot optimise with what it has. So take your own cost per booked trial — whatever it genuinely is — and multiply by 50. That is the weekly spend at which one ad set has full information. For a single-site studio the honest answer is usually a fraction of it, and that is the reason the funnel above has three ad sets rather than nine. A studio optimising for pack purchases, or an access gym optimising for online joins, is closer to the line than any other local business in this cluster, because the event is a checkout and not a phone call.

Then leave it alone. Pausing an ad set, or changing its optimisation event, audience or creative, is a significant edit and restarts learning; budget and bid changes can too, depending on size. An account adjusted every morning is an account that never finishes learning, and in a trade where the owner is also the head coach the daily fiddle is the most expensive habit available.

Now the calendar, which matters more here than in any trade on this site except apparel. January is the category's whole year for a lot of gyms, and it is also the most crowded auction of the year, because every gym within your radius has the same idea in the same week. The price-lock ad on the Browse page is the alternative in one image: sell December to the people who would otherwise join in January, at a rate that locks, and you have moved your cheapest joins out of your most expensive month. September is the second peak — the post-summer restart — and the summer dip is real for the access gym and less so for a studio with a midday timetable. Budget should follow that curve rather than sit flat: heavier in the month before each peak, where the auction is quieter and the reader is already deciding, lighter in the peak itself.

Judge the account on attended first sessions, not on leads — and not on joins either, if your cancellation is genuinely two taps. A lead that never turns up cost you a spot somebody else wanted; a January join that cancels in February cost you the acquisition and a refund conversation. The number that tells you whether the ads are working is how many people are still on the floor in week six, by ad, and that is a spreadsheet the front desk can keep.

The first three are the same mistake at three layers — letting the ad be about the reader's body — and between them they account for most of the rejections and most of the wasted spend in the category.