Two national med spa systems sell the same treatments to the same customers in the same country, and their public ad footprints look almost nothing alike. One runs thousands of ads on Google, overwhelmingly Maps placements. The other runs a few hundred on each platform and leans to the Meta feed. Neither is misconfigured. They are answering the same question, where does a treatment actually get chosen, and arriving at opposite answers. That makes platform mix in a local-service category a strategy fork rather than a setting you copy from whoever is biggest.
This expands on a post I published on LinkedIn about the split between the two largest US med spa systems.
Key Takeaways
- As captured on 13 August 2026, the two largest US med spa systems by clinic count run very different public ad footprints: roughly 9,000 active Google ads at one, against roughly 200 at the other.
- On Meta the gap narrows and reverses. Roughly 180 active ads at the Google-heavy system, roughly 250 at the other.
- Filtering the 9,000 to Maps placements barely moves the count, so that footprint is a bet on map inventory specifically, not on Google generally.
- A map ad and a feed ad address different demand states. One reaches someone ready to book nearby; the other reaches someone who has not chosen a provider yet.
- These are counts of active creatives on the public web, not spend. They tell you where inventory is being bought, not how much it costs or what it returns.
- The practical question is not which system is right. It is which fork your own category sits on, and whether your reporting is even able to tell you.
- Most multi-location reporting rolls placement detail up into "Google" and "Meta," which is exactly the level at which this distinction disappears.
What the Two Largest Systems Actually Buy
The numbers below come from our own reading of the public ad libraries, captured on 13 August 2026 from a signed-in session. Full methodology, advertiser identifiers for independent re-capture, and the earlier captures that show how these readings drift sit in our multi-location healthcare research for the med spa and aesthetics segment.
Milan Laser, which reports more than 400 clinics, showed roughly 9,000 active ads in the Google Ads Transparency Center. Filtering that view to Maps placements barely moves the count. On the Meta Ad Library the same brand showed roughly 180 active US ads.
LaserAway, which reports 226 locations, showed roughly 200 active ads on Google and roughly 250 on Meta.
So the comparison is thousands of map placements at one system against a few hundred ads spread across both platforms at the other. Read it as two very different theories of where a treatment decision happens, not as one operator outspending another.
Two cautions that decide whether you can use any of this. First, these are counts of active creatives, not spend, not impressions, and not revenue. A brand can run one high-budget ad or five hundred small ones. Second, these readings drift week to week and have to be read on their own date: the same two brands read differently in late July than in mid-August, which is why every figure here carries its capture date. Anyone quoting an ad-library count without a date is quoting a number that has already changed.
Why Both Bets Are Rational
The instinct is to decide which one is wrong. That instinct is the mistake, because the two placements are not competing versions of the same ad. They intercept demand at different moments.
A map placement reaches someone who has already decided they want the treatment and is now choosing where to get it. Intent is high, the radius is small, and the competitive set is whoever else appears on that map. Volume scales with locations, which is why a system with hundreds of clinics can rationally hold thousands of active map ads at once: the unit of the campaign is the location, not the brand.
A feed placement reaches someone who has not decided anything yet. It sells the category and the brand before the person is in market, and its unit is the audience rather than the location. A few hundred well-produced creatives can cover that job nationally, because you are not buying presence in every neighborhood, you are buying attention in a demographic.
Both are answers to "where does the choice actually get made." If you believe the decision happens at the moment of local search, you buy the map, and your ad count scales with your footprint. If you believe the decision is made earlier, in the feed, and the local search is just the paperwork afterward, you buy the feed, and your ad count scales with your creative pipeline. The strategies diverge because the underlying belief diverges, and both beliefs are defensible in the same category.
That is also why importing another operator's mix wholesale is a bad idea even when they are larger than you. Their mix encodes their belief, their footprint, and their creative capacity. Copying the output without holding the same belief just buys inventory that does not match how your customers actually decide.
Talk to an Improvado expert about seeing placement-level performance across every location.
How to Tell Which Fork You Are On
This is answerable with data you already have, and it does not require a competitive study.
Look at where conversions start, not where they finish. If most bookings trace back to a branded or near-location search, the decision is happening at the map and your budget belongs there. If bookings trace back to people who first met you somewhere else and searched your name later, you are paying map inventory for demand the feed already created, and the map is taking credit for it.
Check whether your locations behave the same way. They usually do not. A dense urban location with ten competitors inside two miles lives or dies on map presence. A location with no nearby competitor may need the feed to create demand that does not exist yet. One national mix applied to both is two wrong answers. This is the same failure mode behind budget pacing that looks healthy at the network level and is off the rails per location.
Test the fork instead of arguing about it. Hold a set of comparable locations, shift mix meaningfully in one group, and watch booked appointments rather than clicks. The answer is usually visible in a quarter, and it is specific to your category and footprint rather than borrowed from a competitor's ad library.
Watch for the reading to change. A competitor's mix is a snapshot of their current belief, and beliefs get revised. The two systems above read differently three weeks apart. Treat any single capture as a datapoint with a date on it, not as a standing fact about how the category works.
Why Most Reporting Cannot Answer This
The reason this question stays open at most multi-location companies is not that nobody thought of it. It is that the reporting layer flattens exactly the detail the question needs.
Placement is usually the first thing lost. Spend arrives grouped as "Google" and "Meta," sometimes by campaign, rarely by placement type, so the difference between a map placement and a search placement inside the same account is invisible in the roll-up. At that level the strategic fork this article is about simply does not appear.
Location is the second thing lost. Even where placement survives, results are often reported nationally, so a mix that is right for forty locations and wrong for four hundred still shows an acceptable blended number. And the outcome that matters, a booked and attended appointment, usually lives in a different system than the ad platforms, which is why so many multi-location teams end up optimizing to leads rather than to revenue. We have written about that gap in healthcare marketing attribution and about the structural version of it in multi-location healthcare analytics.
Putting placement, location, and outcome on one governed model is the whole job. That is what we build at Improvado: ad platform data joined per location, placement detail preserved rather than flattened, and consistent metric definitions across brands and regions, so a question like "which fork are we on" is a query rather than a project.
Talk to an Improvado expert about joining placement, location, and booked revenue in one model.
Frequently Asked Questions
What does "platform mix is a strategy fork" mean?
It means the split of budget across platforms in a local-service category encodes a belief about where the customer actually decides, rather than a best practice you can copy. Two operators can look at the same category and reach opposite conclusions: one concludes the decision happens at the moment of local search and buys map inventory that scales with its number of locations, the other concludes the decision happens earlier in the feed and buys a smaller set of creatives that scale with its audience. Both are internally consistent, so the useful question is which belief matches your own customers rather than which operator is bigger.
What exactly did the ad libraries show?
As captured on 13 August 2026 from a signed-in session, Milan Laser showed roughly 9,000 active ads in the Google Ads Transparency Center, with the Maps-filtered view barely reducing that count, and roughly 180 active US ads in the Meta Ad Library. LaserAway showed roughly 200 active ads on Google and roughly 250 on Meta. Milan reports more than 400 clinics and LaserAway reports 226, per company announcements. The full methodology and the advertiser identifiers needed to re-capture these readings independently are published with our multi-location healthcare research.
Do these ad counts tell me who spends more?
No, and treating them that way is the most common misreading. They count active creatives visible in public ad libraries, which says where inventory is being bought and in what variety, not what it costs or what it returns. One brand can run a small number of heavily funded ads while another runs many small ones. The counts are useful as evidence of strategy and useless as evidence of budget.
How stable are ad-library readings?
Not very, which is why every figure here is dated. The same two brands read materially differently between late July and mid-August 2026. There is also a methodology trap on Google: a signed-out viewer can be shown an age-gated floor rather than a real count for some advertisers, so a reading taken without signing in can understate a competitor's activity by an order of magnitude. Always record the capture date and the conditions, and re-capture before reusing a number.
Does this only apply to med spas?
No. The mechanism is generic to any category where a customer chooses both a brand and a nearby location: dental, veterinary, urgent care, physical therapy, fitness, home services. Wherever demand can be intercepted either at the moment of local search or earlier in the feed, the same fork exists and the same evidence is available, because Google and Meta both publish ad libraries for every advertiser. Med spa is a useful example because the two largest operators diverge so sharply.
Where do I start if my reporting cannot show placement by location?
Start by getting placement detail out of the platforms before it is aggregated, and keep the location dimension attached to it. Then join whatever your true outcome is, usually a booked and attended appointment rather than a form fill, so that mix decisions are judged on revenue rather than on leads. Until those three things sit on one model with consistent definitions, any answer about platform mix is an opinion, and the fork stays undecided by default rather than on purpose.
