Nutrisense Hit $3.3M a Month on 50% Margins, Not 85%
Nutrisense tripled to $3.3 million a month by selling glucose monitors it doesn't manufacture. The founder's answer on gross margin breaks the usual SaaS arithmetic in half.
On this page
Nathan Latka spends most of his interviews establishing that a company’s gross margin is somewhere in the mid-eighties. Twenty minutes into this one he tried the usual move on Nutrisense and got stopped.
I’m used to interviewing pure SaaS founders, margins are always like 85 percent gross margin, right? I’m getting this sense very quickly — if someone pays you $2,200 for a yearly plan, your margin is not 85 percent.
Nathan Latka, interviewing Dan Zavorotny
“No, it’s not,” said Dan Zavorotny. “No. This is why this makes it fun.” Pushed for the real figure, he landed on roughly 50 percent — and added that when they started, “it was really, really poor.”
Nutrisense is in a SaaS dataset without being a SaaS company. It is direct-to-consumer, it ships hardware it does not make, and half of every subscription dollar goes back out to a sensor manufacturer, a prescribing doctor and a dietitian before anyone at Nutrisense gets paid.
The number in the headline
At the time of the interview Nutrisense was doing $3.3 million a month in revenue, up from just under $1 million a month a year earlier. Zavorotny, asked where they were twelve months before, said “about three and a half less— somewhere around a little under a million a month.”
Those two monthly figures are exactly what the GetLatka profile carries, annualised: $11,880,000 recorded for 2022 and $39,600,000 for 2023. The data is the founder’s own arithmetic, twelve months of it.
Worth flagging: the video’s title claims growth “from $12m to $45m ARR,” and $45 million never appears in the conversation. Twelve times $3.3 million is $39.6 million. Take the monthly number.
What $2,200 a year actually buys
The product is a continuous glucose monitor — a small white sensor that sits on the back of your arm, penetrates the skin a couple of millimetres, and reads your blood in real time — plus an app, plus a dietitian.
Nutrisense makes none of the hardware. “There’s Medtronic, there’s Eversense, there’s Abbott, there’s Dexcom. So we partner with these companies. They manufacture it. What we do really is we focus on the software analytics on top of it.” Each sensor lasts fourteen days, which means two per customer per month, and Nutrisense pays its supplier $40 to $60 per device.
$40–$60 per sensor, two sensors a month, paid to a third-party manufacturer. “These things are eating up a good amount of margin for us.”
A third-party provider writes the prescription. Nutrisense pays for that consult regardless of the outcome — “we have to pay for that no matter what the doctor decides, whether they want a prescription or not.”
Against roughly $225 a month in customer revenue, that is a structurally different business from software. Latka’s reaction on tape — “my gosh, this is complicated” — was the honest one. Zavorotny’s answer was that the complication is the point: “This is why we went into this, because it’s a problem that nobody wanted to attack, because it’s a hard problem to solve.”
The retention curve that made it work
A 50 percent gross margin only survives if customers stay, and Nutrisense’s retention numbers moved by an order of magnitude.
3 weeks → 7–8 monthsaverage customer lifetime, start of the business to the interview
“When we first started, people would stay for like three weeks,” Zavorotny said. “And now we’ve iterated to the point people are staying seven, eight months on average.” The mechanism he describes is unfashionably manual: “We basically call our customer monthly to understand what is it you like. And if we’re close to solving your problem, great — what’s the next problem you want to solve?”
He also refuses to treat lapsed customers as churned by default. Nutrisense sorts people into buckets — one-time users, continuous users, and what he calls health check-ins — and computes lifetime value separately for each. “A lot of people, they don’t need us all the time. They look at us almost like an annual physical. They’re gonna come back year after year for one month, but they’re gonna come back three years straight.”
The governing rule is a ratio, not a duration. “As long as there is at least three to one contribution margin lifetime value to our CAC, we’re always happy with that.” With a weighted average customer acquisition cost around $200 and payback inside two to three months, the ratio holds even on half margins.
The $500 landing page
Nutrisense had revenue before it had a product, and before it had a dollar of funding.
- Find the people already obsessed. Two Facebook groups — the ketogenic diet community and Oura Ring owners. “Those two groups seemed to be really into this right away.”
- Ask for money with nothing built. A landing page, and a price: $500. “We said, pay us and then we’ll let you know when we’re ready. We didn’t even tell them when we’re gonna launch.”
- Let the prepayments be the proof. Zavorotny’s co-founder, an engineer, wanted evidence before writing code. “As I went and got people to prepay, he’s like, whoa, this is real demand. People want to prepay for something that doesn’t exist yet.”
- Ship in three weeks. He “drank a lot of coffee and just sat there and pumped out the first version of the software.”
The launch collided with an iOS release that introduced night mode, blacking the screen from 10pm to 6am. “So people are paying $500 and they can use the app for half, a third of the day. That was a really fun challenge to solve.”
Prepayment stayed as a funding mechanism far longer than the product needed it. “That’s how we didn’t have to raise much money, because we’re able to just get cash and run the business that way.” They removed it later, deliberately trading the cash-flow advantage for a lower barrier: “We want to slowly work toward making us affordable to more people.”
The funding nobody planned
The first round was $250,000 from Techstars and one angel, closed “about three weeks before the pandemic.” The plan after that was to raise millions. “Unfortunately COVID hit and we’re like, alright, guess we have to survive on revenue instead.”
- Sep 2019 · Launch Prepaid landing page, first software build in three weeks.
- Early 2020 · $250k Techstars plus one angel, roughly 10 percent of the company, weeks before lockdown.
- Late 2021 · $1.2M Safes, the round Zavorotny is willing to call a seed.
- 2021–22 · ~$5M A Series A, on standard terms.
- Aug 2022 · $25M The round that takes the total to roughly $32 million.
Fifty thousand people have paid Nutrisense for at least one month. Around 15,000 to 16,000 are active in any given month, and at least 100,000 sensors have shipped — “significantly more than that,” by Zavorotny’s reckoning. The team was 170 people, about 25 of them engineers including data science, and 18 in marketing.
One correction worth making while we are here: the video calls Zavorotny the CEO. He is not. He is the co-founder and COO, and he says so twice on tape — most memorably when asked which CEO he studies.
My co-founder, my CEO. Because my job as chief operating officer, I have to basically be the opposite of him. Whatever he gets good at, I have to go do something else. So I’m constantly just looking at what he’s improving at, what he’s not focused on, so I can complement him in that direction.
Dan Zavorotny, co-founder and COO, Nutrisense
Asked what he wished he had known at 20, he gave the answer of someone who charged strangers $500 for software that did not exist: “Having a mindset that failure is part of the process. It’s great not to fail, but worst comes to worst, you learn a bunch. And there’s no risk, really. What’s the worst that happens? You get a job. That was a fear that I’ve always had, and I never took that chance — but I finally took the chance as an adult.”
Sources — Dan Zavorotny interviewed by Nathan Latka, recorded 22 March 2023. Revenue, headcount, customer and funding figures from the Nutrisense profile on GetLatka, with dates as recorded.
