A lookalike audience lets Meta find people who resemble your best members, then shows your ads to them. For a gym, that means building the audience from a list of real, paying, active members and aiming it at the neighborhoods you can actually serve. Done well, it replaces broad "fitness interest" targeting with something much closer to the person who joins and stays. This guide covers which lists to use, how to prepare them, how to keep the audience local, what the law expects and how to tell whether it worked.
TL;DR
- A lookalike audience is built from a seed list. The better the seed (retained, paying members), the better the prospects Meta finds.
- Meta needs at least 100 people from one country in the seed, but a list of 1,000 or more tends to give steadier results.
- Lookalikes are built at the country level, so you keep them local with location targeting on the ad set, not in the audience itself.
- Get consent before you upload member data. The FTC treats buried privacy-policy language as insufficient for sensitive uses.
- Judge the campaign on cost per booked visit and cost per new member, not on clicks or likes.
Table of Contents
- What a lookalike audience is, and why gyms use it
- Choosing your seed list
- Preparing the list before you upload
- Building the lookalike in Meta Ads Manager
- Keeping a lookalike local
- Privacy, consent and Meta's rules
- Testing lookalikes against other audiences
- Measuring whether it worked
- Our perspective
- When to hand it off
- FAQ
- Sources
What a lookalike audience is, and why gyms use it
A lookalike audience is an advertising segment built by an algorithm. You hand the platform a seed audience of existing customers, and it finds other users who share demographics, interests and behaviors with that seed. In plain terms: you give Meta a list of people you wish you had more of, and it goes looking for them.
For gyms the appeal is specific. Selecting "fitness" as an interest captures everyone who ever liked a workout page. A lookalike built from paying members can instead find people who resemble those members and are likely to join, rather than people who are merely interested in fitness. One lead-quality guide warns that default interest targeting often produces reflex form-fillers with low purchase intent, which is the pattern most owners describe when they say "Facebook leads never show up."
A lookalike is not magic. It copies patterns in your seed. If the seed is a pile of one-time trial signups who never came back, Meta will find more people like that. The work is in choosing the seed.
Choosing your seed list
The seed is the single biggest lever. Start from the question: who do I want ten more of?
Seeds that tend to work
- Active, paying members. One gym software guide suggests members who attend at least five times per month as a starting point.
- Long-tenured members. Another source points to members active for six months or longer, which filters out the January crowd that quits by March.
- Members on your best plan. If your unlimited or coached tier is the profitable one, seed from it.
- Engaged audiences as a fallback. If you lack a big list, Meta also lets you build custom audiences from email lists, website visitors and people who engaged with your Page or Instagram. These are weaker signals than paying members, but better than nothing.
Seeds to avoid
- Everyone who ever filled out a form, including no-shows.
- Expired free-trial leads who never converted.
- Members who joined on a deep discount and left within a month.
- A mixed list of staff, friends, vendors and members.
How big should the seed be?
Meta's floor is low: the source must contain at least 100 people from a single country. The same guide notes Meta recommends a source of roughly 1,000 to 5,000 people, while stressing that quality matters as much as size. A small CrossFit box may have 150 to 300 active members, so you are likely near the floor. Two ways to work around it:
- Combine current active members with former members who stayed a long time and left on good terms (moved away, injury, schedule).
- Add website visitors or lead-form engagers to widen the base, and test it against the member-only version.
A small seed is not a reason to skip lookalikes. It is a reason to test them against a simple local interest audience rather than assume they win.
Preparing the list before you upload
Bad formatting wastes your seed. Before upload:
- Export from your member software as a CSV with email and phone columns.
- Clean it. Remove duplicates, staff, test accounts and anyone who asked not to be contacted.
- Standardize the fields. One practitioner guide recommends lowercase emails, no stray special characters and phone numbers in E.164 format, and aims for a match rate above 70%. Treat that as a rule of thumb, since it comes from one marketer's experience. A low match rate in Ads Manager still tells you something: the list is stale or poorly formatted.
- Check accuracy and currency. Source data should be accurate, up to date and compliant with privacy rules. An old list full of people who moved is a weak teacher.
If you can, add a column for tenure or plan so you can build more than one seed (for example, "members over 6 months" versus "all active").
Building the lookalike in Meta Ads Manager
The flow is short:
- In Ads Manager, open Audiences and create a Custom Audience from your customer list. Upload the CSV and map the columns.
- Wait for the upload to process and check the match rate.
- Select the custom audience and choose Create a Lookalike Audience, as one gym software walkthrough describes.
- Choose the country and the audience size. Percentages run from 1 to 10, where a smaller number is a tighter match and a larger one is broader.
- Name it clearly, such as "LAL 1% | Active 6mo+ members | US", so you can read results later.
For gyms, a 1% to 3% lookalike is a common starting range. Start at 1% and add a 2% to 3% version only if the first one stops delivering at a healthy cost. Meta's interface changes often, so if a menu label differs from what's above, follow the current prompts in Ads Manager.
Keeping a lookalike local
A gym lives and dies by drive time. A lookalike that finds your ideal member 400 miles away is worthless. The catch: a lookalike is created for a country (or region), not for a city. The local part comes from the ad set.
Here is how we keep it tight:
- Set the ad set location to your service area. Use a radius or a list of ZIP codes and towns around the gym. The lookalike then picks people who resemble your seed, but only within that footprint.
- Size the radius to how people actually travel. Look at your member list by ZIP code. If 80% of members live within a short drive, your radius should reflect that, not the largest circle Meta allows.
- Use "people living in this location" rather than people recently in it, so day visitors and tourists are not served your ads. (For paid search, one gym ads guide covers the same location-setting logic.)
- Watch for a small pool. In a rural or small-town market, a tight radius plus a 1% lookalike can be too small for Meta to deliver efficiently. Widen the lookalike percentage before you widen the radius.
- Exclude your own members from the ad set so you are not paying to advertise to people already on the roster. Customer lists make this simple, since you can exclude existing customers from acquisition campaigns.
The trade-off is real. A tight local radius limits scale. If a campaign cannot spend its budget at a decent cost, that is the signal to expand the area or loosen the lookalike, not to add more interests.
Privacy, consent and Meta's rules
Member lists are personal data, and gyms often hold more sensitive information than they realize: attendance patterns, injuries, health goals, body measurements. Treat that with care.
- Consent matters most for sensitive data. The FTC says businesses must obtain affirmative express consent before using information collected in a confidential context for advertising or another purpose the person did not request. It also says that burying consent terms in a privacy policy or terms of service is not clear and conspicuous.
- Enforcement is real. The FTC took action against a treatment provider over alleged disclosure of health data to advertising platforms without consent. That case involved addiction treatment, not a gym, but it shows how regulators view health-adjacent data shared with ad platforms.
What that means in practice: upload only contact details (email, phone) rather than health or workout data, make sure your signup flow tells members plainly that their contact info may be used for advertising, and honor opt-outs. Keep health and injury notes out of any file you send to an ad platform. State laws and Meta's own terms for customer lists vary and change, so have your attorney review your waiver and signup language, and read Meta's current terms for custom audiences before you upload.
Testing lookalikes against other audiences
Do not assume a lookalike wins. Run a clean test.
| Audience | What it is | Best used for |
|---|---|---|
| Lookalike 1% (member seed) | Strangers resembling your best members | Cold acquisition in your radius |
| Lookalike 2%-3% | Broader version of the same | Scaling once 1% saturates |
| Local broad targeting | Location only, no interests | Letting the creative do the filtering |
| Interest stack (fitness, CrossFit) | People who follow related pages | Baseline comparison |
| Retargeting | Site visitors, video viewers, form-engagers | Warm follow-up |
A workable approach: put the same offer and creative into each cold audience, give each equal budget, and compare after at least two weeks. One agency guide suggests keeping roughly 70% of the budget on cold acquisition and 30% on warm audiences, adjusting after about two weeks of cost-per-lead data, and holding off on warm expansion until the campaign has run four to six weeks with at least $50 a day in total budget. Those are one source's suggestions, not rules, so treat them as a sane starting point. For the warm side, see our guide to paid social retargeting for gym owners.
Change one variable at a time. If you swap audience and offer together, you will never know what moved the numbers. Your offer matters as much as the audience; our piece on how gym offers drive more leads and memberships covers what to put in front of a cold lookalike.
Measuring whether it worked
Meta will happily report reach and likes. Neither pays your rent. Track cost per lead, cost per booking, show-up rate and cost per new member instead.
A worked example with made-up numbers, for illustration only: two audiences each spend $500. The interest audience brings 25 leads ($20 each), 8 bookings and 2 members. The lookalike brings 18 leads ($27.78 each), 9 bookings and 4 members. The lookalike looks worse on cost per lead and better on cost per member ($125 versus $250). That is why you follow the lead all the way to the sale.
A few more measures:
- Lead quality. One source defines high-quality gym leads as people who reply, confirm interest, have the means to join and convert at 20% or higher. Use it as a benchmark to aim at, not a promise.
- Frequency. Some experts consider a frequency of 1 to 2 optimal. In a small local radius you will see it climb fast, which is a cue to refresh creative.
- Speed to lead. A great audience still fails if nobody calls back within minutes. Read how to convert more CrossFit gym leads into paying members.
None of this works without tracking that connects ad clicks to members. Our guide to conversion tracking for gyms shows how to count members, not clicks, and why tracking fitness ad performance matters before you scale spend.
Our perspective
We work only with gyms and fitness brands, and the same pattern shows up again and again: owners blame the audience when the real problem is what happens after the click. A lookalike is a good tool, but it is a multiplier, not a fix. If your offer is vague, your landing page is slow, or leads sit for a day before anyone replies, a better audience simply delivers more people into the same leak.
So we sequence the work. First the offer and the follow-up. Then tracking that ties a new member back to the ad. Then the audience tests. In that order, a 1% lookalike from your best members becomes something you can trust and scale. In the wrong order, you spend a month wondering why "the audience didn't work."
We also lean on a rule of thumb for small boxes: do not wait for a perfect seed. Start with the best list you honestly have, run it against a plain local audience, and let your own member numbers decide. For more on this, see our overview of lead generation mistakes to avoid for fitness gyms.
— Collin
When to hand it off
Building audiences, cleaning lists, testing creative and reading the numbers takes hours every week. If you would rather spend them coaching, our paid advertising service for gyms on Meta and Google handles the setup, testing and reporting. Our plans start at $1,500 per month for Get Found, $2,500 per month for Fill the Floor and $4,500 per month for Growth Engine, and you can compare them on our pricing page.
FAQ
How many members do I need to build a lookalike audience?
Meta requires at least 100 people from one country in the source audience. Meta recommends about 1,000 to 5,000, but a smaller gym can still test with its best active members. Just compare the result against plain local targeting before committing budget.
Should I use a 1% or a 10% lookalike for a gym?
Start at 1%, the tightest match, since gyms are limited to a local radius anyway. Move toward 2% or 3% if the audience is too small to spend your budget or your costs climb. A 10% audience is rarely a good first choice for a local business.
Can I upload my members' emails to Meta?
You can use customer lists, but you need proper consent and clear signup language. The FTC expects affirmative express consent for certain uses of information collected in confidence, and says burying it in a privacy policy is not enough. Upload contact details only, never health data, and ask your attorney to review your waiver.
How long before I know if a lookalike is working?
Give it at least two weeks of steady spend, then judge it on booked visits and new members, not just cost per lead. A full sales cycle for a gym is often longer than the ad's learning period, so check membership starts too.
Sources
- Lookalike Audiences: What They Are and How To Use Them
- Facebook Ads for Gyms: Targeting, Creative & Budget
- Why Your Gym Leads Are Low Quality (And How to Fix It)
- 3 Simple Ways to Boost Member Numbers
- Facebook Ads Targeting Options
- What is a Lookalike (LAL) Audience?
- Lookalike Audiences in Meta Ads: Complete 2026 Guide
- Meta Ads Lookalike Audiences: A Complete Guide
- How Do You Use Customer Lists to Create Powerful Meta Audiences
- FTC Notice of Penalty Offenses Concerning Misuse of Information
- FTC press release on health data shared with advertising platforms
- Meta Ads for Personal Training Packages at Your Gym
- How to Create a Facebook Ad for Your Fitness Business