Should Your Medical Practice A/B Test Its Ads?

A med spa owner called us a while back, frustrated. She had been running the same Facebook ad for four months. It worked okay at first, then slowly got more expensive. Her question was fair: "Is this just how ads go now, or am I doing something wrong?" We asked how many versions of the ad she had running. There was a pause. "One. The one my cousin made." That was the whole problem, and also the easiest kind of problem to fix.

Running a single ad is the marketing version of ordering the same dish every time without ever glancing at the rest of the menu. Maybe it is the best choice. You will never know, because you never tried anything else. A/B testing is just trying something else on purpose, in a way that tells you which option actually brings in patients for less.

What A/B testing actually means

Strip away the jargon and it is simple. You run two ads that are identical except for one thing. Same audience, same budget, same offer, but version A has one photo and version B has a different photo. The ad platform shows each to similar people, and after a while one version is clearly getting patients to click and book for a lower cost. You keep the winner, and you retire the loser. Then you test the winner against a new challenger. That is the whole loop.

The one rule that makes it work: change one thing at a time. If version A and version B have different images, different headlines, and different offers, and B wins, you have learned nothing you can use. Was it the photo? The words? The deal? You cannot tell. Change one variable, learn one lesson, repeat. Boring and disciplined beats clever and random here.

The number that makes the case

This is not a nice to have. According to Meta's own ad testing guidance, winning A/B tests delivered about 30 percent lower cost per result on average. Read that again with your own budget in mind. If you are spending 2,000 dollars a month on ads, a 30 percent improvement is the difference between 40 new leads and 57, for the exact same spend. You did not add a dollar. You just stopped funding the weaker ad.

And healthcare leads are not cheap to begin with. In WordStream's 2025 Facebook Ads benchmarks, the average cost per lead in the health and medical category was about 41 dollars, well above the cross industry average of roughly 28 dollars. When every lead costs that much, shaving the price down is one of the highest return things you can do with an afternoon.

What to test first (in order)

You do not test everything at once, and you do not test the trivial stuff first. Work top down, biggest lever to smallest:

• The image or video. This is the single biggest factor, because the visual is what makes a thumb stop scrolling. A warm photo of your real team usually beats a glossy stock image. Test a real face against a result, or a photo against a short video.
• The first line of the headline. Once you have attention, the opening words decide if they keep reading. Test an outcome ("Walk in without back pain") against a question ("Still putting off that knee?") against a plain offer ("New patient exam, 49 dollars").
• The offer itself. A free consultation versus a flat price versus a limited time slot can pull completely different numbers. This one surprises people the most.
• The audience. Same ad, two different groups. You may find patients in one zip code or age range book twice as readily as another.

Notice what is not at the top of that list: button colors, tiny wording changes, the exact font. Those are last resort tweaks once the big things are settled. On a practice sized budget, chasing them first is a waste.

How long to run it before you call a winner

Here is where most people trip. They launch two ads on Monday, peek Tuesday morning, see version A ahead, and shut off version B. Then they wonder why the "winner" fizzles. Early numbers lie. On tiny samples, one lucky click can make a loser look like a champion.

Give a test at least a full week, so it runs across both weekdays and a weekend, and wait until each version has gathered a few dozen clicks or leads before you trust the gap. If the two are neck and neck after that, it often means the thing you tested does not matter much to patients, which is itself useful to know. Move on and test something bigger.

Does this work for Google Ads too?

It does, and Google automates a chunk of it. With responsive search ads, you load several headlines and descriptions, and Google mixes and matches combinations, leaning toward the ones that perform. You can also run a proper experiment that splits traffic between two ad versions, or between two landing pages, and reports which one wins. Different buttons, same instinct: give the system real options, measure what books patients, and cut what does not earn its place.

The part everyone forgets: test where the ad sends them

You can win the ad and still lose the patient. If your best performing ad sends people to a slow, confusing page, you paid for the click and then watched them leave. The page the ad points to is part of the test. A fast page that lets someone book in two taps will out convert a pretty page that makes them hunt for a phone number every time.

This is why we treat ads and the destination as one system, not two separate projects. A great ad and a website that actually converts have to be built to work together, or you are optimizing one half of a handoff and losing the other.

Why practices skip testing (and why that is costing them)

Three reasons, usually. They are busy running a practice, not a marketing department. They think testing needs a data scientist and a spreadsheet. Or they got one ad working once and are scared to touch it. All three are understandable, and all three are expensive. The cost does not show up as a bill. It shows up as the patients you could have booked for the same money and did not. That is the worst kind of waste, because it is invisible. You never see the patient who scrolled past the ad that was never tested against a better one.

Testing is not a one time project either. What wins in January can tire out by April, the way that med spa owner's single ad slowly got pricier. Patients see the same creative too many times and stop reacting. A steady habit of feeding in fresh challengers is what keeps your cost per patient from drifting up month after month.

How EtherealMinds runs it

We only do healthcare, so we are not guessing about what moves patients in this space. When we manage ads for a practice, testing is not a special event, it is the default. We launch new campaigns with multiple versions from day one, hold everything constant except the one thing we are measuring, and let the data, not anyone's opinion, pick the winner. Then we roll the winner forward against a new idea and keep going.

All of it plugs into the wider patient acquisition system, so a winning ad hands off to a fast page, an easy booking flow, and follow up that catches the people who do not book on the first visit. The ad getting cheaper is only worth something if the patient actually makes it to your schedule. If you want the full picture on what each new patient is costing you right now, we broke that down in our guide to a good cost per lead for a medical practice.

That med spa owner? We put her cousin's ad up against two new versions with real photos of her staff and a clearer offer. Within three weeks one of the challengers was booking consultations for about a third less than the old faithful. Same budget. More chairs filled. The only thing that changed was that she stopped running one ad and hoping.