There's a man in my neighborhood who gets walked by his dog. Ninety pounds of Labrador, leash stretched tight, and the dog picks the route every morning. Left at the corner for a squirrel, across the street for a smell, full stop at the hydrant because another dog stopped there yesterday. He holds the leash, so technically he's the owner. Anyone watching knows who's running the operation. He calls it a walk. Frankly, it's a negotiation he loses before he finishes his coffee.
I think about that man every time an operator tells me their social media plan is to watch what's trending and jump on it.
Related: Leveraging Social Media
Here's the uncomfortable part. Social media is shaping what patients ask for, and that isn't going away. The 22-year-old who walks in asking for the treatment she saw on TikTok is the new normal; the trend cycle now sets the consultation before you've said a word. So the question for an operator isn't whether the algorithm influences your patients. It does. The question is whether you're holding the leash or getting dragged down the block by it.
An algorithm is a trainable animal. It is not a slot machine, and that distinction is the whole ballgame.
Most med spas treat their feed like a slot machine. Pull the lever, post the trending sound, hope for the jackpot. Maybe this one goes viral; maybe the next one does. It's gambling dressed up as marketing, and like all gambling, the house has already done the math on you. A trained animal is different. You give it consistent signals, the same commands, day after day, and it learns what you want and brings it back. Feed it junk, and it learns to fetch junk. Feed it the right signals, and it learns to bring you the right patients.
Related: Social Media Savvy: Jennifer Levine, M.D.
That reframe changes three things about how you run the channel.
First, the algorithm learns from what you feed it, so feed it on purpose. Every post is a command, whether you meant it as one or not. Chase a viral dance, and you've told the machine to bring you people who like viral dances; you'll get reach, and most of that reach will be tire-kickers, teenagers, and the discount crowd shopping ten med spas for the lowest price on a syringe. Publish consistently about the work you actually want to do, for the patients you actually want to serve, and the machine slowly learns to put you in front of them. A good trainer doesn't suppress the dog's energy and instinct; he aims it. Same idea here. You're not fighting the algorithm's reach; you're pointing it.
Second, virality is the jackpot fantasy, and fantasies make poor business plans. The metric that pays rent is not views; it's the right patient finding you on purpose, then booking, then coming back. Reach without fit is noise that costs nothing to make and earns nothing to keep. Run the falsification test on it. If going viral built premium practices, the most-followed aesthetic account in your market would be the most profitable one in your market. By and large, it isn't, and often it's the reverse, because the operator chasing the jackpot is burning real hours producing content that trains the machine to deliver the wrong room. A full waiting room of the wrong patients isn't a win; it's a more expensive problem.
Third, and most importantly, the human scrolling the feed isn't even the real prize. The machines reading your content are. Search engines, and increasingly the AI tools your patients now ask for recommendations, are indexing everything you publish and quietly deciding when to put your name forward. That's the channel that compounds. A trending video dies in 48 hours; a body of consistent, credible, well-tagged content keeps working while you sleep, teaching the crawlers who you are and who you serve. We run zero paid ads at our practice, and that's not a budget decision; it's a conviction that a durable channel teaches the machines the truth about your business, not renting attention by the click. Paid reach stops the second you stop paying. A trained channel keeps fetching.
Picture two practices posting the same number of times a week. The first one reacts: trending audio on Monday, a meme on Wednesday, a rushed take on whatever a celebrity did on Friday. The second one repeats: a plain explanation of one signature treatment, an honest before-and-after, a straight answer to a question patients actually ask. A year later, the first has a feed full of dead trends and a camera roll of vanity metrics. The second has a library the search engines and the AI tools can read, sort, and recommend, because it says the same true things a hundred different ways: same effort, opposite compounding. One practice trains the machine; the other gets trained.
None of this means ignoring trends. A trend is information; it tells you what's on your patients' minds this month, and that's worth knowing. The discipline is in the response. You let the trend inform the conversation without letting it rewrite your menu. When a patient walks in asking for the treatment of the week, that's your opening to do the thing the algorithm can't: read the room, ask what she's actually trying to fix, and guide her to the right answer instead of the popular one. The machine brought her to the door. What happens after the door is still yours, and it always will be; the algorithm doesn't have kids who go to the same school as your patients' kids.
So here's the commander's intent. Stop treating your content like a lottery ticket and start treating it like training. Pick three signals you want the machine to learn about your practice: the treatments you want to be known for, the patient you're built to serve, the standard you refuse to drop. Publish against those three on a steady cadence for 90 days, no chasing, no detours for whatever's loud that week. Then judge the results on the only number that matters: the quality of the consultations walking through your door, not the view count on your best post.
The man with the Labrador could fix his mornings in two weeks with a trainer and a little consistency. He hasn't, because being dragged is easier than learning to lead, right up until the dog runs into traffic. The operators who win the next few years online won't be the ones who went viral. They'll be the ones who decided, early, that they were going to train the channel instead of letting the channel train them.
Hold the leash. Give the commands. The algorithm is going to learn something from you either way; the only question is whether you taught it on purpose.
