What you actually need to start a telehealth practice

Ask this question anywhere and you will get a list. An EMR. A CRM. A scheduling tool. A pixel. A CDP. A telehealth platform with a patient portal. Possibly an agency retainer to run ads into all of it.

The list is not wrong, exactly. Every item on it is a real thing you will eventually need. The list is wrong about when, and when is the only part of the question that costs you anything.

Here is the pattern from four months of calls with physician founders, and it is remarkably consistent: the expensive EMR and the expensive CRM are already bought. The patients are not there yet. And the founder does not describe this as a mistake. They describe it as being prepared.

The instinct is trained

This is the part that generic startup advice misses, because generic startup advice is not written by physicians. Doctors do not overbuy because they are careless with money. They overbuy because they were trained to a standard of evidence that makes overbuying feel like diligence.

"We want to overprepare for everything. For example, when we do clinical studies, we set it up so we'd only get fooled by chance less than 5% of the time, so that's a pretty high bar. Something that I think people carry across the different fields is that we're very conservative, which you should be when you have people's lives in your hands. Marketing is a lot less high stakes, but people want to be overprepared, and that's the problem."

Pranay Parikh, MD

The convention he is pointing at is real. In clinical research, a result is conventionally treated as significant below a p-value of 0.05, a threshold set so that if there were no real effect you would rarely see a result that large by chance. Whatever you think of the convention, notice what it trains. It trains you to assume you are wrong until the evidence is overwhelming. It trains you to build for the bad case. In medicine that is not pedantry, it is the job. The cost of being wrong is measured in people.

Then you start a company, and the training comes with you.

The payoff structure is inverted, and nobody tells you

In the clinic, the asymmetry runs one way: being wrong is catastrophic, being slow is usually survivable. You can order another test. You can bring the patient back in two weeks.

In a new business the asymmetry runs the other way, and this is the whole thing. Being wrong about a tool costs you a subscription. Being slow costs you the business. You cannot bring the market back in two weeks.

So the same conservatism that makes you a good doctor makes you buy a CRM for the patients you do not have yet. The instinct is not defective. It is correctly calibrated to a different game.

"Why do you need a CRM when you don't have any patients? CRM is an email software... email them from your support email until you get a couple hundred patients."

And the bleed is not one big decision you could catch. It is small enough, every time, to feel reasonable:

"When you're starting a new business it's like death by a million paper cuts. It's like $100 here, $200 there, $300 there, and we pay it because we make good money, and then all of a sudden you're like, man, I'm losing $2,000 a month on this."

The phrase we make good money is doing quiet work in that sentence. A physician founder can absorb each individual charge without noticing. That is precisely why the total arrives as a surprise.

The cost side is what kills things, even with infinite money

If you think the answer is simply to spend more and get there faster, consider what happened to the largest retailer on earth when it tried exactly that.

Walmart launched health centers in 2019. In April 2024 it closed all of them, and shut the virtual care arm too. In its own announcement: "Through our experience managing Walmart Health centers and Walmart Health Virtual Care, we determined there is not a sustainable business model for us to continue." The reason it gave was not that nobody came. It was this: "the challenging reimbursement environment and escalating operating costs create a lack of profitability that make the care business unsustainable for us at this time."

That is a company reporting $648 billion in revenue that fiscal year. Money was never the constraint. Escalating operating costs were.

Be careful with the analogy, because it only carries one thing. Walmart ran fifty-one physical centers, and reimbursement is a payer problem that most cash-pay telehealth founders do not have. A $648 billion retailer and a solo practice do not fail the same way. But the mechanism is the mechanism: the cost structure got them, at a scale where the budget was infinite. Death by a million paper cuts is not a small-company disease. It is just faster at small companies.

Which should end the argument that pouring money in makes it work sooner. It did not work for the people who had all of it.

The order that actually matters

None of the above is an argument for buying nothing. It is an argument that the sequence is the product, and almost everyone runs it backwards.

Positioning comes first, and it is not a branding exercise. It is the answer to why a patient picks you over the funded competitor who can outbid you in the auction tomorrow. Until that answer exists, every dollar downstream amplifies a message that does not land. In Pranay's words: "We shouldn't be talking about paid ads at all. Because that is just going to be money wasted before your positioning is figured out."

Organic starts immediately, not because it is cheap, but because it is slow. This is the piece founders reverse most often, treating organic as the thing you do when you cannot afford ads. It is the opposite. It is the thing you start first precisely because it compounds slowest, and because it is the only channel that tells you what to say before you pay to say it:

"it takes the longest. So you want to get that started as soon as possible. And two, you can start to understand what kind of stuff people like to hear about. And then start driving ads to it."

Small tests come next, and their job is to find out, not to grow. Five or ten dollars a day buys information, not patients. Founders who skip this step are not being aggressive; they are paying full price for the lesson.

Real spend comes last, on a message that has already proven it works on people who were not paid to see it.

Underneath all four is the rule that generalises past tools entirely:

"I would also recommend doing everything manually at first, but then from that manual process, we look at it after a couple months. Then see what we can automate."

Manual first. Automate what survives. The CRM question is not really about CRMs. It is one instance of that rule. It applies identically to hiring, to your intake flow, to the EMR, to every piece of software that promises to handle a volume you have not reached. Automating a process you have never run by hand means paying to scale something you do not yet understand.

What this actually takes

It would be dishonest to end here with a checklist, because the hard part of this was never the list.

The hard part is that the sequence feels wrong while you are doing it. Starting lean feels like under-preparing, which, if you were trained the way you were trained, feels like negligence. Doing intake manually feels like you are not serious. Spending ten dollars a day feels like you are not really trying. Watching a competitor with a real budget launch a real stack while you email patients from a support address feels, on a bad afternoon, like proof they are ahead of you.

They may not be. They may simply be further along a more expensive path to the same discovery.

None of that is a weekend fix. You cannot buy your way out of this and you cannot checklist your way out of it either, because the work is unlearning a conservatism that took years to install and that still serves you well in the room where it belongs. Most people do not do it. They keep preparing for a rush, and the rush arrives, when it arrives, as a trickle, into a business that has been paying for the rush all along.

The instinct that makes you careful with a patient is the one costing you here. It is a good instinct. It is just in the wrong room.