Hims has 2,442 employees. Medvi has two and better margins.

In partnership with

A New York Times piece dropped on April 2 that I keep coming back to.

Matthew Gallagher launched Medvi, a direct-to-consumer telehealth provider for GLP-1 drugs, out of his Los Angeles house in September 2024. He had $20,000 in starting capital. No co-founder. No investors. More than a dozen AI tools.

In its first full year the company posted $401M in sales, signed up 250,000+ customers, ran a 16.2% net margin, and is projected to do $1.8B in 2026.

The team is two people. Matt and his younger brother Elliot. They're doing roughly $3M a day.

Now look at Hims & Hers. Same product category. $2.4B in revenue last year. 2,442 employees. 5.5% net margin.

The math is brutal in a way Silicon Valley has not fully processed yet.

Three things I haven't been able to stop chewing on.

Lesson 1 — The org chart is the moat now

Gallagher used ChatGPT, Claude, Grok, and Replit to write his code, generate website copy, produce ad creative, build customer service flows, and run business analytics.

Quote from the NYT piece: "It's not an A.I. company, but I did it with A.I."

That sentence matters more than the revenue number.

Most of the AI conversation last year was about whether AI products would replace SaaS products. Medvi is the first time the answer landed at the operations layer instead of the product layer. The product is a generic telehealth flow that's existed for a decade. The leverage is in the org chart.

Functions that used to need humans, code, copy, creative, support, analytics, dashboards, are now one person and a stack of tools. The cost structure compounds in his favor for as long as the tools keep improving. Hims has to pay 2,442 salaries through that same window.

The era of "build the team, then build the company" is over for a specific category of business. We're going to see a lot of these in the next 18 months.

Your clients can't tell you dictated that.

That polished proposal? Dictated in two minutes. The detailed follow-up email? Spoken between meetings. The Slack update with clean formatting and exact numbers? Voice.

Wispr Flow turns everything you say into professional, send-ready text. Speak naturally — go on tangents, change your mind, circle back — and get output that reads like you spent twenty minutes writing it.

89% of messages sent with zero edits. Millions of professionals use Flow daily. Works in every app on Mac, Windows, and iPhone.

Start flowing free

Lesson 2 — Boring product. Specific founder. Better margins.

Notice what Medvi did not invent.

It didn't invent GLP-1s. It didn't invent telehealth. It didn't invent direct-to-consumer pricing. Hims, Ro, and a dozen others established the playbook a decade ago.

The wedge is purely operational. Faster. Cheaper. Two people instead of two thousand.

That's the entry pattern most founders this year are going to miss. People keep looking for a category to invent. The win is sitting inside categories that already work, run by 2,000-person operators with single-digit margins, waiting for somebody to rebuild them with a tenth of the headcount.

If you're staring at an industry you understand and you've been telling yourself "someone already does this," that's the signal, not the disqualifier. The question isn't whether someone does it. It's whether they do it with 2,442 people.

Most do.

Lesson 3 — The backstory you've been hiding is the wedge

This is the part most readers will skim and shouldn't.

Gallagher is 41. He grew up in a trailer park. Before Medvi, he ran a watch subscription company called Watch Gang that did $300M in sales. He's a digital marketer by training, not a doctor and not an engineer.

By the standard founder playbook he has the wrong resume. By the actual playbook he has the only resume that matters.

Twenty years of buying performance ads, running DTC funnels, and managing subscription cohorts is exactly the muscle you need to take a saturated category like telehealth and ship a faster version with a smaller team.

He didn't pivot to telehealth. He redeployed direct-response marketing into a new product wrapper.

The version of this story most NTE readers haven't given themselves permission to consider is the second-act version. Gallagher built his $300M watch business in his 30s, then went again. The second one is going to be 6x bigger and built by two people.

The founder backstory the standard narrative tells you to hide is almost always the actual wedge.

One thing the NYT piece is honest about

Medvi's customer service bot hallucinated drug prices in the early months. Gallagher honored every fake price the bot quoted. The bot also invented product lines that didn't exist, and he had to correct those by hand.

The FDA sent Medvi a warning letter in February about some of its on-site claims.

Neither of those derails the lesson. They reinforce it. The two-person company means the founder is the only human backstop when the AI does something stupid or the regulator notices. That's the trade.

If you're building a Medvi-shaped business, you're signing up to be on call for every AI failure mode, every angry customer, every FDA inquiry. The leverage is real. So is the personal load. Most founders glossed over that part of the NYT piece. Don't.

Where this connects to what we do at NTE

The reason NTE Pro exists is because the Medvi pattern is the dominant founder pattern of the next 24 months. You don't need a new category. You need an existing category with a fat-margin incumbent and a workflow you can run with two people and an AI stack.

The 7,000+ ideas in NTE Pro aren't there for you to copy. They're there to train your eye until you walk into your gym, your kid's pediatric dentist, or your mortgage broker's office and within ten minutes you have three Medvi-shaped opportunities mapped in your head. The muscle is buildable.

For the operators tracking the next Gallaghers before the NYT writes them up, WhoFiled is the live feed of state filings, trademark drops, hiring patterns, and small-team signals. The next two-person company doing $3M a day is sitting in this morning's data. Find them now, not in April.

Your weekend assignment

Pull up the largest company in your industry on LinkedIn. Count their employees.

Now imagine running 80% of their revenue with two people. Where does it break? What do you actually need humans for? What can a Claude or a Grok or a Replit absorb that you'd been assuming required a team?

Hit reply with what came up. The headcount number. The function you couldn't replace. The function you suddenly realized you could. I read every reply.

Tomorrow morning I'll send the five-step check I'd run on myself before lunch if I were trying to find my own Medvi-shaped business inside an industry I already know.