How to split your budget between Google Ads and Meta Ads for a gym

For a local gym spending under €1,000 a month, the short answer is: most of the money goes to Meta, a smaller slice to Google, and never 50/50. The 50/50 split is what you set when you don't want to think, and it's almost never optimal. This article explains the logic behind the split, gives concrete allocations by spend level, and, more importantly, shows you how to measure which channel is actually bringing in real members, because that's where almost everyone gets it wrong.

The two channels capture different moments

Before talking percentages, you need to understand what each euro buys on each platform, because they don't buy the same thing.

Google Ads captures active demand. Someone types "gym near me" or "crossfit in [city]" into the search bar. That person has already decided they want to join something; they're just choosing where. That's the hottest lead this business generates. The problem is there aren't many of them. And because there are few of them and every gym in the area is bidding on them, they're expensive: clicks from €1.50 to €4 depending on city and niche.

Meta Ads captures latent demand. Your ad interrupts someone who was scrolling through their brother-in-law's videos on Instagram. That person wasn't looking for a gym, but they live 800 metres from yours, they're 35, and they've been telling themselves for three months that they really should do something. The volume is enormous, the cost per lead is low (typically €4–12, see CPL benchmarks for gyms), and the temperature is cold: many of those leads will need conversation, follow-up, and time.

Operational summary: Google gives you few, expensive, hot leads. Meta gives you many, cheap, cold leads. The full comparison is in Facebook Ads vs Google Ads for gyms, but to decide the split, that sentence is enough.

Why the default answer is Meta first

Here's the opinion, and it's a firm one: for a local gym, Meta is the primary channel and Google is the complement. Not the other way around. The reason is arithmetic, not ideology.

Active gym-related demand in your area is finite and small. In a Spanish city of 100,000 people, the total monthly relevant searches ("gym near me", "gym + neighbourhood", "join a gym", variants) typically runs between 1,500 and 3,000 searches a month. That sounds reasonable until you do the maths: with an ad CTR of 5–8% and sharing impressions with the three or four competitors also bidding, you're left with maybe 60–120 capturable clicks a month. At €2.50 per click that's €150–300 in spend and the inventory is gone. You can't put €800 into local Google Search because there aren't €800 worth of searches to buy; Google will spend it anyway, but on progressively broader and weaker match types.

Meta doesn't have that ceiling at gym scale. The 25–55 age group within a 3 km radius is tens of thousands of people, and the algorithm can spend €500, €1,000, or €2,000 a month finding the receptive ones without the cost degrading much. That's why Meta is the scalable channel and Google is the capture channel.

What exactly does Google capture with its smaller share? Three things, in this order:

1. Brand search. People typing your gym's name. Extremely costly to lose (a competitor can bid on your brand) and extremely cheap to defend: €20–50 a month.
2. "Near me" and high-intent local terms. The core of active demand.
3. Display and YouTube remarketing to people who already visited your website. Cheap and reasonably effective as a reminder.

How to set that up properly is covered in the Google Ads guide for gyms. And if you're tempted to let Google make all the decisions on its own, read our take on Performance Max for a local business first.

Recommended splits by spend level

Two notes on the table. First: Google's percentage grows with the budget not because Google gets better, but because Meta covers the essentials first and marginal money performs better diversified. Second: these percentages are starting points, not dogma. After three months, your own data takes over.

The exceptions that change the split

A newly opened gym with no brand presence or reviews: even more Meta, possibly 100%. Nobody searches for you by name because nobody knows you, and in generic local search you're competing against gyms with 200 reviews when you have 4. Meta doesn't penalise being new; Google does, in practice.

Niche with strong search volume: CrossFit, reformer Pilates, martial arts. Here people do actively search ("CrossFit box Seville", "Pilates machine near me") and those searches have real volume with less generalist competition. A CrossFit box can justify a 50/50 or even Google-first split in its early months, because it's capturing demand that formed on its own.

Budget of €300 or less: one channel only, and it's Meta. Splitting €300 across two platforms guarantees neither one exits the learning phase.

How to measure which channel actually wins

Measurement mistake number one: comparing channels by CPL. The Google lead at €25 looks like a disaster next to the Meta lead at €8. Until you look at conversion to member: active-search leads convert 2–3 times better than cold Meta leads, because they'd already made up their mind. A 20% close rate versus 7–8% is typical.

The only metric that compares channels honestly is cost per member:

By CPL, Meta wins easily. By cost per member, they're basically tied. And this example is representative: when both channels are running well, cost-per-member numbers tend to converge much more than the CPL gap suggests. If one channel doubles the other in cost per member over a full quarter, that's a signal to move budget.

To do this you need to trace every signup back to its source channel. At gym volumes (5–15 signups a month), you don't need attribution software: a "channel" column in your leads spreadsheet and the discipline to fill it in.

The last-click trap

There's a flaw in the measurement above worth naming: last-click attribution systematically favours Google.

The typical sequence goes like this. Someone sees your Instagram ad three times over two weeks. They don't click any of them. In the third week, on a Tuesday, they decide they're serious, open Google, type your gym's name (or "gym + your neighbourhood"), click your Search ad, and fill in the form. Last-click attribution gives the member to Google. But Meta manufactured that member; Google just took the basket to the checkout.

Meta plants the seed, Google harvests it. You see this empirically when you pause Meta for a month: branded searches for your gym drop 20–40% within a few weeks, and Google's "performance" drops with them. That's why you should never evaluate your brand campaign in Google as if it were independent acquisition: much of it is the harvest of what Meta planted.

This doesn't invalidate cost per member as a metric; it qualifies it. In practice: treat Google's brand campaign as infrastructure (you pay for it and that's that, it's €20–50), and compare only Meta against Google on generic terms. And be sceptical of any report, yours or an agency's, that uses last-click to justify moving budget from Meta to Google. Fine-tuning your Meta campaigns has far more upside than that transfer.

How often to review the split

Quarterly. Not weekly, not monthly. Quarterly.

The reason is statistical: with 8–12 signups a month split across two channels, monthly samples are noise. One month Google "wins" because you closed 3 out of 11 leads; the next it "loses" because you closed 1 out of 12. Neither means anything. You need 25–40 cumulative signups per channel before a cost-per-member comparison says something, and that's three months for most gyms.

What you do review weekly is the health of each channel individually: CPL spiking, high frequency, junk search terms leaking into Google. That's maintenance. Moving budget between channels is strategy, and strategy based on one week of data is astrology.

A sensible schedule: split review in January, April, July, and October. Move a maximum of 10–15 percentage points per review. If the data is calling for a bigger change, it's probably not the split that's broken, it's the execution of one of the channels, and you fix that within the channel, not by cutting its budget.

The split on one page

If you run a normal gym in a normal city: start with 80/20 in Meta's favour, protect your brand on Google from day one, measure cost per member and not CPL, mentally discount the brand harvest, and review calmly every quarter. The exceptions (niche with strong search, no-brand gym, minimum budget) are listed above and there aren't many.

The hard part isn't deciding the split. It's executing both channels well simultaneously with the time you have left after actually running a gym, which is usually none. That's exactly the problem Pilotium was built for: the Meta campaign, the Google campaign, and the follow-up on every lead all run on their own and you see members, not dashboards. But with or without a tool, the principle doesn't change: the complete acquisition plan relies on both channels doing their own job, not competing for the same euro.

One check for this week: search your gym's name on Google from your mobile. If a competitor's ad is sitting above your listing, your first euro of Google spend already knows where it needs to go.