Superwall’s Joseph Choi talks to Asya Polony, CPO of Welltory — the Apple Watch app that reads physiological signals to show where your energy drains. 51 minutes, $18M ARR, ~18M downloads, and a full walkthrough of her team’s onboarding deck.

The six-ingredient pitch formula

The framework Asya uses for onboarding (and says works for pitching anything):

  • Human problem — get specific and emotional about when the user actually feels the pain (“you’re on the couch doom-scrolling TikTok again”). She spent months listening to a wellness podcast on repeat because the host was the user. AI can’t do this part for you.
  • Product promise — bold but curiosity-sized, not the full solution: “Your Apple Watch knows what makes you tired and stressed.”
  • The key — insider knowledge nobody else has: sedentary stress (your heart works hard while you sit still). “Today you already have 192 minutes of it” — pulled from their real data.
  • Mechanism — show how the app delivers: real-time visualization of when bad stress happens and when you relieve it.
  • Proof — the strongest proof is showing users their own data live; reviews, millions of users, and scientific journal publications are supporting cast.
  • FOMO — name the bad thing that happens if they don’t pay today: “stress is building right now, and you won’t see what’s driving it.”

Onboarding decisions that break the rules

  • First CTA is “sync your health data” — no intro slides, no value pitch first. Over 90% of users hook up their data. Every time she tried a “prelude,” it performed worse.
  • Notifications asked immediately — against every best-practice guide; it works best for them anyway.
  • The buy-in screen is about emotion, not logic — “Do you ever feel tired even when you haven’t done much all day?” puts people back in the feeling, then the data lands.
  • Empowerment before the close — after showing the 30-day stress calendar (“you’ve only relieved ~20–30% of it”), pivot to “your body was also recovering — here’s the timeline, we’ll show you when.”
  • 60%+ of conversions happen after day one — a long tail, unusual for health apps; the product keeps showing people their own data until the habit and value compound.

Paywall psychology

  • $99/yr (billed annually; sometimes a monthly option) — mostly annual-only, hard-ish paywall, no trial.
  • Benefits tested worse than fear: “you’ll have more energy, feel great” underperformed “stress is building right now — don’t you want to see what’s driving it?”
  • Category rule: benefit-selling works for trigger apps (Uber, DoorDash — the pain is already immediate); fear/FOMO works for long-horizon categories like health and language learning where the pain isn’t felt at the moment of onboarding.
  • Free vs paid: everything vital and potentially life-saving (blood pressure, “you’re about to get sick” signals, basic metrics) stays free; the deeper analysis (trends, timelines, artery-elasticity recommendations) is paid.

The AI take

  • “Less data and more narrative” — Welltory’s AI direction is making data tell the story of how your body lived the day, layered with what the user says they were doing, rather than dumping numbers into a chatbot.
  • Her north star: an AI that knows how your body feels at all times and adapts its tone accordingly — “the way we experience the world is informed by what’s happening inside our bodies.”
  • Pulling the user’s own device data is the personalization hack — cheaper and more magical than gen-AI personalization from survey answers.

“Stress is not bad. What’s bad is too much of it left unrelieved over long periods of time.”