Social Media for Clinicians

On this page

You can build a real audience as a clinician without betting your license to do it. This guide is the playbook I use and teach: what to post, what it costs, what to hand to other people and to AI, and the lines you do not cross. It is written for licensed clinicians — physicians, NPs, PAs, and colleagues in adjacent licenses — and especially for the ones building something of their own. If you run a practice or a healthspace, or want to, your audience is not a side project. It is the pipeline that feeds the thing you own.

90 days
Do it yourself first
$150–500
Per video, the first hire
~1 year
Until it compounds

Why are you posting?

Decide why you are posting before you touch a camera. The reason picks your platform, your format, and the metric that tells you whether it is working — and skipping this step is how clinicians end up 18 months into a content habit that serves no goal.

Five reasons cover almost everyone I have advised:

  • Teaching at scale. The explanation you give four times a day in clinic could reach forty thousand people instead. Watch saves and shares, not follower count.
  • Growing a practice. You want patients or members who already trust you before the first visit. Watch booked consults that mention your content.
  • Opening doors. Speaking slots, collaborations, advisory work. Watch who shows up in your inbox.
  • Advocacy. You want to move how people think about a clinical issue. Watch how often your framing gets borrowed by others.
  • Building a name while employed. You are planting a flag for the practice you have not founded yet. Watch your email list, because it is the only asset that leaves with you.

If you are building a practice or a healthspace of your own, be honest with yourself that the audience is the pipeline. That is not cynical. It is the reason to hold the work to a clinical standard.

Which platform should a clinician start with?

Start on the platform whose output compounds toward your reason for posting — and for most clinician-founders that means YouTube plus an email list, with everything else as distribution. Here is the map I actually use:

PlatformBest forEffortWhat it converts to
YouTubeDurable search presence; depthHigh per pieceHighest-intent viewers: consults, sign-ups
LinkedInReaching decision-makersModerateReferrals, partnerships, B2B doors
InstagramReach and familiarityModerate, steady cadenceAwareness that warms other channels
TikTokFast discoveryHigh cadenceTop-of-funnel attention
Email listAn audience you ownLow, compoundingEverything — it survives every algorithm change

One more filter before you commit: match the format to what you are already good at. A natural explainer should film. A sharp writer should start where writing wins — LinkedIn and a newsletter. If you think in visuals, build around diagrams and carousels. Copying another clinician's format because it worked for them is the fastest way to become a worse version of someone else, and the compounding never starts.

Should you do it yourself or hire someone?

Do it yourself for the first 90 days. Not because help is bad, but because you cannot supervise work you have never done, and in this field you are accountable for every word published under your name.

The sequence I give every clinician who asks:

  1. Write your reason for posting in one sentence. Tape it somewhere visible.
  2. Pick one format on one platform. One.
  3. Make ten pieces yourself, end to end — topic, script, recording, edit, caption. My first ten were bad, and I have deleted them since. That was tuition, not waste. Yours will be bad too, and that is the point of doing them cheaply.
  4. While you make them, study three to five peer accounts. Choose them by criteria, not fame: same specialty or same audience, one to two years ahead of you, and an output level you could actually sustain. You are studying their systems — cadence, structure, how they handle comments — not their charisma.
  5. Only then hire, and only for tasks you have personally done and can judge.

The clinicians who skip to step five fund an expensive experiment in discovering what they should have learned in steps one through four. Agencies will happily run that experiment for you.

What does help cost?

Anywhere from intern rates to five figures a month — and the freelance editor is the highest-leverage first hire almost every time. Rough market ranges:

HelpTypical cost, as of mid-2025When it makes sense
Marketing intern$15–22/hourRepeatable tasks you have already systematized
Freelance video editor$150–500 per videoFirst hire — buys back your highest-cost hours
PR or full-service agency$3,500–30,000/monthRarely at the start; only with proven output and a specific ask

The editor math is simple: an hour of your clinical time is worth more than an hour of cutting silences. The agency math is usually backwards: you pay the most money at the moment you have the least ability to evaluate the work.

What does a sustainable weekly system look like?

A fixed order you repeat until it is boring. I run the same sequence every week, and the fixed order is the reason it survives a clinical schedule:

The weekly engine: which stages stay yoursFive stages in order: research is delegable, narrative stays yours, batch recording is your time, editing is delegable, and final approval stays yours.ResearchBatch recordEditNarrativeApproveAI drafts, you verifyclaims and orderone sittingeditor cutsyour name on it
Solid stages never leave you. Outlined stages are the delegable middle.
  • Research. Collect the week's questions — from patients, comments, colleagues — and pull the evidence. This is the stage I now delegate to AI first drafts and verify myself.
  • Narrative. Decide the one point each piece makes and the order it unfolds in. This stays mine. It is the actual clinical work of the whole system.
  • Batch record. One block, phone or camera on a tripod, several pieces in a sitting. Batching is what makes the schedule survivable; recording one video five times a week is how people quit.
  • Hand off the cut. The editor takes the raw files and returns drafts. I approve every final cut, because my name is on it.

Keep the delegation map explicit: topic choice, claims, and final approval never leave you. Everything else is negotiable.

Where does AI fit?

Hand AI the middle of the pipeline and keep the judgment.

That single rule sorts nearly every use case.

Where it earns its keep for me: synthesizing literature into a first-pass summary I then verify against the sources, drafting graphics, cleaning transcripts, cutting silences. The Health AI Toolkit catalogs the specific tools and prompts I actually use for this.

Where it has no business being: deciding what you claim, standing in as the source of credibility, or publishing a clinical statement no licensed human reviewed. Your audience is trusting a clinician. The moment the judgment is outsourced, that trust is unearned — and unlike an edit, it does not come back with a correction.

How long until this works?

Plan on a year of compounding before it feels like it is working, and treat anyone promising faster as a red flag. The honest timeline:

PeriodWhat to expectWhat it costs
Months 0–3Near-zero metrics; you are learning the craft$80–150 for a mic and a light
Months 3–6Search traffic begins; tens to hundreds of subscribersYour time, mostly
Months 6–12Thousands of followers; first consults that mention your contentA first hire, if output is steady
Year 2Referrals from content become routineA system, not a sprint

The curve is flat and then it is not. Every clinician I know who built a durable audience describes the same shape — and almost quit in the flat part.

What are the rules?

The rules are not fine print. They are the reason you still have a license and an audience in year five, so treat them as load-bearing:

  • Patient storiesDe-identify past self-recognition, or get written consent. Strip and swap every identifying detail that is not doing the teaching — age, timeline, specialty context, the memorable oddity — until the person in the story could not find themselves in it. If the details are the lesson, ask in writing before you post.
  • Teach, don't treatTeach principles in public; practice medicine in visits. The moment a question turns case-specific — their labs, their meds, their symptoms — the answer is an appointment, not a reply. Say so kindly and move on.
  • Stay inside the evidenceAnd say when it is thin. Certainty you do not have reads as authority right up until it costs you credibility, or a board complaint. "We do not know yet" is a complete sentence and audiences respect it.
  • Disclose every financial tieIn the content itself — not in a bio, not behind a link, but in the post or video where the endorsement happens. The FTC's endorsement rules apply to clinicians with no carve-out for good intentions.
  • Read your contract firstMany employment and partnership agreements claim approval rights, restrict topics, or require disclaimers. Finding that clause after you have an audience is a much worse conversation.

None of this is physician-only. NPs and PAs answer to their own boards and collaborative-practice terms, and those boards read social media the same way medical boards do.

How do you own your audience?

An email list — it is the only channel no algorithm change can take from you. Platform followers are rented attention on infrastructure someone else controls; the platforms have repeatedly reminded creators of this, and clinicians are not exempt.

So every platform you post on should have one job beyond its native metric: moving people onto the list. For a founder this is not a nice-to-have. The list is the front door of your practice — the place where an audience member becomes a patient, a member, or a referral source.

Followers are reach. The list is equity.

How would I decide today?

If I were starting from zero this week, knowing what I know now: I would write my reason for posting in one line and let it pick the platform. I would choose the single format closest to what I already do well, and make ten pieces alone before spending a dollar on help. I would start the email list on day one, not day ninety — everything I published would point to it. When output was steady enough to hurt, my first hire would be a freelance editor, for the hours I am worst at and mind the most. And at 90 days I would sit down with honest numbers and my one-line reason, and decide whether to double down, adjust the format, or stop on purpose rather than fade out.

Common questions

Do I need my employer's permission to post?

Check before your first post, not after someone forwards one to your medical director. Many employment and partnership agreements claim review rights over public statements, restrict named-employer commentary, or require disclaimers. Knowing the terms early is cheap; renegotiating them with an audience already built is not.

Can I talk about my patients?

Only if the story is de-identified past the point where the patient could recognize themselves — or with their written consent. Changing the name is not de-identification; the details that make a story compelling are usually the ones that identify. When in doubt, teach the principle without the story.

Should I hire an agency to get started?

No. Make ten pieces yourself first. You will learn what the work actually is, what good output looks like, and which single task is worth paying for — usually editing. Hiring before that knowledge exists means paying the most at the moment you can evaluate the least.

Does this playbook change for NPs and PAs?

The mechanics do not change at all — platforms, systems, costs, and timelines are identical. What changes is which rulebook binds you: your own board's social media expectations plus any collaborative-practice or supervision terms, which can add employer-style approval clauses of their own.

Sources

Building a practice of your own? See how CareCore works with clinicians.