Zach Yadegari

Cal AI did $30M in 2025, then $5.7M in a single January

Zach Yadegari's calorie-tracking app Cal AI finished 2025 at $30M in revenue and posted $5.7M in revenue in January alone, implying a roughly $50M+ annual run rate before its sale to MyFitnessPal.

“I could tell you that right before selling, we finished 2025 having done $30 million in revenue. And now the company's grown. Um, in January, we just did $5.7 million in revenue.”

From an unblocked-games site to a $100K exit at 16

Zach learned to code at 7 to build video games, then as a high school freshman built an unblocked-gaming website for classmates that grew to 5 million users (mostly via his own TikToks) and $60K/yr in banner ads, which he sold at 16 for $100,000 before moving into apps.

“I built an unblocked gaming website that lets students play games in school Unblocked. It was actually kind of accidental that I stumbled into turning it into a real company. First, I just built it for my friends at school, but it spread really quickly and it grew to 5 million users through mostly me just making TikToks about it. It was generating $60,000 a year by putting banner ads on the website. And then when I was 16, I sold it for $100,000.”

Use the Pareto principle to learn enough to hire people smarter than you

Zach's rule for moving fast across domains: learn the 20% that gets 80% of the results, which is enough to hire people smarter than you in that area and manage them effectively.

“I always think that the most important thing is the Pareto principle. Learn the 20% that gets you the 80% of results. And that's what's necessary to actually hire someone that's smarter than you and being able to manage the team.”

Steal thisLearn just the 20% of a skill that gives 80% of the value, then hire experts you can now credibly manage.

Get warm intros to buyers framed as the buyer's own idea

Zach was told cold outreach loses you leverage in a sale process, so he had people he knew make warm intros to target acquirers, framing the meeting as if the connector thought of it rather than as the founder asking.

“you lose leverage if you're just cold outreaching. So I made, I tried to get people that I knew to make warm intros, kind of saying something along the lines of like, hey, I think you guys should meet. This is a cool company. I just thought of this like as if it was his idea, right? Not if it was me asking him.”

Steal thisSource acquisition conversations through warm intros framed as the connector's idea, not your cold ask.

Hire an M&A banker to surface buyers you've never heard of

A founder told Zach how M&A bankers actually work: they bring a list of acquirers you've never heard of, often companies about to go public that want to roll up smaller companies, and will pay fair market value for your business whenever you want to sell.

“he basically explained how bankers work. M&A bankers, you go to them, they create a list of all these companies you've never even heard of before. And like these companies that are about to go public, they just want to roll up all these smaller companies into them. And so they're willing to give you actually fair market value for your company anytime you want.”

Steal thisBefore deciding your company is unsellable, engage an M&A banker who can surface roll-up buyers you'd never find yourself.

Cal AI's exit started as a 'freemium intel' call with MyFitnessPal

Planning to make Cal AI freemium for longevity, Zach booked a call with MyFitnessPal's CEO to mine him for freemium tactics under the guise of catching up; the CEO shared nothing, but the conversation pivoted to a possible partnership and ultimately the acquisition.

“So I wanted to hop on a call with their CEO and squeeze whatever information about freemium I could out of him, pitch it as like catching up, but just try to ask him questions about freemium. Reached out, got on a call, and they did not answer a single question about freemium. And it makes sense. I mean, like, it would definitely be not in their interest at all to help us. But yeah, like, they, they actually spun the conversation from talking about that to, okay, maybe something else makes sense.”

Influencers got Cal AI to $2M/mo, then paid ads broke the plateau

Zach says influencer marketing took Cal AI to about $2M/month before plateauing; the next leg of growth came from investing heavily in performance ads on Instagram, TikTok and Facebook, which now fuel most of the company's growth.

“influencer marketing is what started us and it took us really far. We got to about $2 million a month just from influencer marketing. The thing is, after that, we kind of plateaued a little bit. And what really got us over the hump, the next step was running ads, performance ads. On Instagram and TikTok and Facebook. And so invested a lot more time learning and scaling ads. And that's actually what's fueling most of the growth at this point.”

Steal thisUse influencers to find product-market fit and early scale, then graduate to performance ads to break through the plateau.

Cal AI's $500K MrBeast sponsorship paid off via credibility, not clicks

Zach paid MrBeast $500K to appear in a video and directly recouped only ~$350K-$400K, but argues the deal was profitable long-term because name-dropping MrBeast gave the brand credibility that unlocked many more deals worth more than the original spend.

“We probably made $350,000, maybe $400,000 just from that $500,000 promo. And the rest of it, actually, I believe we were profitable long-term. And if we're not yet, we're going to be because you having the brand name MrBeast saying we were in a video, we were able to do way more deals afterwards. We had way more credibility.”

Use expected value (EV) to decide whether to sell

Zach applies the gambler's expected-value model to selling a company: multiply each outcome's probability by its payoff, including the chance the whole thing falls apart, and pick the path with the higher EV instead of deciding on emotion.

“EV is a good one. So like gamblers obviously know this where it's like the percent chance of winning something and then the prize and you multiply them together. And then it's like whatever you have a 100% chance of getting versus the chance of losing it all versus, versus the prize money you could have, whichever is greater, like that's which one you should often pick.”

Steal thisScore a sell-vs-hold decision by EV: probability times payoff for each path, including a real chance of going to zero.

Semantic search across your own Instagram/X followers

Zach pitches an app that lets you semantically search your own followers on Instagram or X (find the entrepreneurs, find your dating type) and get alerted when someone interesting follows you, calling Instagram one of the biggest de facto dating apps; he frames it as an easy few-million-dollar idea.

“someone should build a semantic search for your followers on either Instagram or X, wherever you want so that you could look through. And this could be helpful for so many reasons. Let's say you're looking to date someone like Instagram. A lot of people don't realize, but like, it's one of the, it's one of the biggest dating apps.”

Steal thisBuild semantic search plus 'interesting person followed you' alerts over a user's existing Instagram/X follower graph.

Quitter: a $5M/yr app that gamifies quitting porn

Zach cites a friend's app, Quitter, that helps men quit porn with no real AI, just streaks, a roadmap, rewards and meditation guides, and is now doing over $5M a year, offered as proof that simple gamified consumer apps remain a great space.

“my friend built an app called Quitter. It is an app that helps guys quit porn. Doesn't really use AI, it just kind of gamifies the process, gives you a streak, gives you a roadmap, rewards, and like meditation guides. And that's been super helpful for a lot of people. And now it's making over $5 million a year.”

Claim: enroll people in class-action lawsuits, $1M in 30 days

Zach cites a friend's app, Claim (spelled CAIM), that lets users sign up for class-action lawsuits to collect $10-$15 settlements; it made about $1M in its first 30 days before being pulled from the App Store.

“It actually got taken off the App Store, unfortunately, but they made like a million dollars in their first 30 days of launching.”

Steal thisProductize 'go find/save me money for free' as an app, like auto-enrolling users into class-action settlements.

'That already exists' is proof a market works, not a reason to quit

Zach argues founders are wrongly deterred when a competitor already exists; a growing incumbent is proof the market works and is big enough that an underserved gap remains, so existing competition should attract you, not scare you off.

“It should say that's proof. Yeah, exactly. That's proof it works. And the market is so big that all of these apps can be successful. I mean, if a company is still growing, that means that there's still a gap. There's still people that are underserved and can use it. Like I hear all the time, I was in the entrepreneurship class, people were pitching their ideas. Everyone was like, wait, that already exists. That already exists. And it like deters them, but it shouldn't.”