How Cal AI Grew to $30M a Year (The Playbook)
The short answer
How did Cal AI grow to over $30M a year?
Cal AI, a photo calorie app, grew past $30 million a year in under two years with five moves: make one action take seconds, test one trusted creator before scaling (one story for about $2,000 brought about 1,000 downloads and about 100 payers), pay creators for views you can count, make ads as ads, and charge at the core action while fixing churn in the product.
Key facts
- Over $30M a yearCal AI's annual revenue when MyFitnessPal announced it had bought the app, with over 15 million downloads and seven employees (MyFitnessPal's figures).TechCrunch, MyFitnessPal has acquired Cal AI, 2 March 2026
- About $2,000The cost of the first influencer story, which brought about 1,000 downloads in two days and about 100 paying users (the founder's account).Net Influencer, Cal AI founder breaks down the influencer playbook (TechCrunch Disrupt 2026 talk), 22 October 2026
- About $5,000 a dayWhere profitable Facebook spend topped out when Cal AI reposted influencer videos as ads; ads made on purpose took revenue from about $2M to $4M a month (the founder's figures).Net Influencer, Cal AI founder breaks down the influencer playbook (TechCrunch Disrupt 2026 talk), 22 October 2026
- Anthropic and OpenAICal AI estimates calories from a meal photo using models from both companies; the founder said different models are better with different foods.TechCrunch, photo calorie app Cal AI downloaded over a million times, 16 March 2025
Chapters
Transcript
Edited for readingOne $2,000 post, then $30M a yearWatch from 0:00
Cal AI, a photo calorie app, started with one influencer story. It cost about $2,000. A thousand downloads in two days, the founder says. Around a hundred of them paid. Two of its founders were still in high school.
In under two years, it passed $30 million a year. Then MyFitnessPal bought it.
Last month, at TechCrunch Disrupt, the founder explained what worked. So here are the five moves behind it, from public interviews and reporting, and what you can copy.
Move 1: one job, done in secondsWatch from 0:30
Cal AI does one job. You take a photo of your meal, and it estimates the calories. That isn't a new idea: MyFitnessPal does a meal scan too. So why did this one take off?
MyFitnessPal's own CEO gave the answer. He said there's an audience of people who want it fast, and AI based. TechCrunch put it simply: Cal AI is for people who prefer speed over accuracy.
Under the hood, it uses models from Anthropic and OpenAI. The founder said different models are better with different foods.
Our take: they didn't win with a new idea. They won by making one action take seconds.
Move 2: test one trusted creatorWatch from 1:07
Before Cal AI, the founders launched six apps in about a year. This time, they found an influencer who was sharing his weight loss journey. They paid him about $2,000 for one story, showing him scanning a meal.
How many downloads do you think that brought in two days? About a thousand. And roughly a hundred of them paid.
The founder credits the influencer's following, not the content: a tight community that trusted him.
Then they signed up influencers as fast as they could. $30,000 in month one. Over $100,000 in month two. By month six, $1 million a month.
Our take: one small paid test told them the plan. Before they spent big, they knew the message worked.
Move 3: pay for views you can countWatch from 1:50
The founder calls mass user-generated content, lots of cheap creator videos, a spray and pray strategy. Influencer marketing, he says, is a sniper approach. That's what really took them off.
They looked for tight communities, not the biggest accounts. The rough minimum was 50,000 followers, but views counted for more. And they wanted creators whose audience wanted to be like them.
Here's the clever part. Small creators often asked flat prices that wouldn't pay back. So Cal AI paid the asking price, but tied it to a view target or a set number of videos.
Our take: pay for results you can count, not one post and a hope.
Move 4: make ads as adsWatch from 2:30
Next, they ran their influencer videos as Facebook ads. It hit a ceiling: profitable spend topped out around $5,000 a day. Videos that did well on their own didn't work as paid ads.
So they recruited creators to make ads on purpose. More direct. More salesy. He says that took revenue from $2 million a month to $4 million.
Our take: a post and an ad have different jobs. Make each one for its job.
Move 5: charge at the core actionWatch from 2:58
Now the money. On the App Store, Cal AI says food scanning needs a subscription. So the one thing everyone downloads it for is the thing you pay for.
They learned this earlier. Their Grind Clock app was free with ads, and it made little money. Moving to a hard paywall multiplied revenue about ten times, the founder says.
And to keep people paying? Beyond email, they ran no retention campaigns. They made the food scan more accurate instead, and retention rose as accuracy improved.
Our take: charge at the moment of value, and fix churn inside the product.
Why MyFitnessPal bought itWatch from 3:34
When the deal was announced, MyFitnessPal said Cal AI had over 15 million downloads, and over $30 million in annual revenue. The founder put it at over $40 million in sales over twelve months. The team? Seven employees, plus a few contractors.
MyFitnessPal tracks about seventy competitors. Cal AI caught its eye early in 2025. They talked for almost a year. The deal closed in December and was announced in March.
The playbookWatch from 4:07
So here's the Cal AI playbook, in five moves. One: make a single action take seconds. Two: test one trusted creator before you scale. Three: pay creators for views you can count. Four: make ads as ads, not reposts. Five: charge at the core action, and fix churn in the product.
The big money came later. The first proof cost about $2,000.
Sources
- TechCrunch, MyFitnessPal has acquired Cal AI (15M+ downloads, $30M+ annual revenue, seven employees, deal timeline; figures from MyFitnessPal) (2 March 2026)
- Net Influencer, Cal AI founder breaks down the influencer playbook (TechCrunch Disrupt 2026 talk; all growth figures are the founder's own) (22 October 2026)
- TechCrunch, photo calorie app Cal AI downloaded over a million times (models from Anthropic and OpenAI) (16 March 2025)
- App Store listing, Cal AI Calorie Tracker (read 3 Nov 2026) (3 November 2026)
Caveat: Phone screens in the video are simplified illustrations, not Cal AI's interface. Revenue and growth figures are as reported by MyFitnessPal and the founder; they are not audited. The founder's over $40 million in sales figure comes from MyFitnessPal's press release of 2 March 2026 on GlobeNewswire.

