The storyHow they got there
Zach Yadegari was 17 years old and sitting in a high school classroom when he noticed the same problem everyone around him had: tracking calories was so painful that nobody did it consistently. MyFitnessPal required typing out every ingredient. Lose It required searching a database. Nobody had cracked the obvious solution — point your phone at food and let AI figure it out.
He built the first version of Cal AI with his co-founder Blake Anderson over a few months, using off-the-shelf vision APIs to do the heavy lifting on calorie estimation. The product was deceptively simple: take a photo, get a calorie count in under three seconds. No accounts required. No database to scroll through. Just a camera and a result.
The insight that made the distribution work wasn't the product itself — it was realizing that food photos are native TikTok content. Every user demo was a potential viral video. The use case was relatable to anyone who'd ever tried to lose weight. And the result appeared on screen fast enough to fit into a 15-second clip.
They seeded the product by reaching out to 50 fitness micro-influencers via DM — creators with 20K to 100K followers who were already posting gym and nutrition content. The pitch was direct: "We built a calorie tracker. No typing. Just take a photo. Want to try?" They offered free premium access plus $500–$2,000 to post once. Several early videos hit over a million views and drove thousands of installs each.
Rather than relying on a single account, they ran what became a 12-account TikTok strategy — separate accounts targeting different angles, demographics, and content formats. Each account was a parallel distribution experiment. When a format performed, they doubled down across all channels simultaneously. When it didn't, they killed it fast.
By month four, they'd hit $1M MRR. By month six, they were ranked #9 in the App Store health category with 500,000 downloads. By month twelve, they'd crossed $30M ARR with a team of fewer than 30 people — most of whom were still in their early twenties.
The paid acquisition layer ran alongside the organic. TikTok Ads, Facebook, and Instagram retargeted people who'd engaged with influencer content. The CAC on blended spend stayed around $3–7 — exceptionally low for a consumer app — because the organic viral coefficient was pulling so much weight. Unit economics were clean: LTV around $140, CAC around $5, a 20:1 ratio that most SaaS companies would trade their entire product roadmap for.
MyFitnessPal CEO Mike Fisher reached out in late spring 2025 after watching the growth trajectory. The deal closed in December 2025 for an estimated $40–50M. Zach was 19 years old. The Cal AI team of 30 was retained and the product continues operating as a standalone brand under the MyFitnessPal umbrella.
The lesson isn't "build a calorie app." It's that in 2024–2025, TikTok plus micro-influencer seeding plus paid performance marketing is the repeatable distribution playbook for consumer apps — and that a product designed to be demoed in 15 seconds on camera has an unfair advantage in that environment.
Channel MixWhere the growth actually came from
Most case studies hand-wave channels. Here's the rough allocation — not in dollars spent, but in users acquired — across the routes that actually mattered.