How Ditch Optimized Onboarding by Scoring Its Screens

Wealth building used to start at $0. For many young Americans, it now starts well below that, with the need to payoff significant student and/or credit card debt.Luis Andino saw this while working as a data scientist at LendingPoint, where he found that many Americans start their adult lives by taking out high interest personal loans to pay for college or buy their first car.In 2023, he left to build Ditch for what he calls the “-1 to 0” phase of wealth building.The first version of Ditch was a testing ground. Luis and his team built a sophisticated platform for experimenting with new features and tracking user behavior before investing significant time and energy into building out their product. Then in May 2024, they pivoted Ditch to automatically roundup everyday purchases and apply this spare change toward debt repayments. Users could now watch their debt go down a little bit every week without lifting a finger. With that change, the product took off.After taking two years to reach $1M in ARR, Ditch took 80 days to get to $2M in ARR, 59 days to hit $3M in ARR, and just 16 days to approach $4M in ARR.In this episode we cover:How Ditch thinks about owning the “-1 to 0” phase of building wealthWhy Ditch spent 2 years strengthening its unit economics before scalingHow Ditch optimized new user onboarding by scoring every screen in its flowDitch’s process for turning organic UGC on TikTok into paid adsWhy Ditch built processing fees and refinancing tools on top of its subscription Lessons Luis has learned raising money in a difficult fundraising environmentLuis Andino:Website: https://ditch.com/ LinkedIn: https://www.linkedin.com/in/luisoscarandino/ X: https://x.com/Iuicho TikTok: https://www.tiktok.com/@notluichoPhil Carter:Website: https://elementalgrowth.com/Substack: https://philgcarter.substack.comLinkedIn: https://linkedin.com/in/philgcarterX: https://x.com/philgcarter

“Sleep is a waste of time.”That was Leon Sasson’s mindset in college until he partnered with classmate Jeff Kahn to start researching sleep with Northwestern’s sports teams and saw what it did for performance.Other teams heard about the results, so Leon and Jeff cofounded Rise as a high-end business offering one-on-one sleep coaching to sports teams at thousands of dollars per athlete per year. After struggling to scale this model, Leon and Jeff pivoted Rise into a consumer app in 2020 with a new mission: bring what they had learned helping elite athletes sleep better to anyone with a smartphone, without the need for any wearable device.Rise is now one of the top sleep apps in the world, used by over 10 million people. In addition to its core sleep tracking business focused on sleep debt, Rise also recently launched an AI coach and a clinical care platform that helps patients with conditions like insomnia and sleep apnea. In this episode, Leon shares:Why Rise built its product around sleep debt to maximize a user’s “energy schedule”How Rise connects new users to the core value of its product in their first 60 secondsHow Rise built a rigorous and high-velocity experimentation cultureWhat the company found when it revisited A/B tests that it had previously called “losers”How Rise thinks about Oura and Whoop in the increasingly competitive sleep spaceHow Rise is expanding beyond sleep tracking into AI sleep coaching and clinical careThe three specific sleep tips he believes matter most for long-term health and wellnessLeon Sasson:Website: https://www.risescience.com/ LinkedIn: https://www.linkedin.com/in/leonsasson/ X: https://x.com/leonsasson Phil Carter:Website: https://elementalgrowth.com/Substack: https://philgcarter.substack.comLinkedIn: https://linkedin.com/in/philgcarterX: https://x.com/philgcarter

“We weren’t giving the user a real reason to upgrade.”That’s how Ajay Mehta, cofounder and president of Tolan, describes the first version of the free tier for his company’s AI companion app.Freemium opened up the top of the funnel, but it resulted in an initial short-term hit to subscriber conversion and ARPU, which was particularly problematic given that every session a user spends with their Tolan results in non-trivial AI costs.So Ajay’s team spent a year running hundreds of A/B tests that boosted subscriber conversion +200% and ARPU +218%, ultimately paving the way to Tolan’s recent announcement that it had raised $30M in additional funding from Union Square Ventures and Khosla Ventures.In this episode, Ajay returns to Subversive as the podcast’s first repeat guest to discuss:Why Tolan shifted its focus from Gen Z women to Millennial momsHow it decided which features to give away vs. put behind the paywallUsing cheaper open source models to make its free tier sustainableThe specific A/B tests that helped Tolan maximize subscriber conversionHow a single paywall test directly drove a 40%+ lift in ARPUWhy the company expanded from one subscription tier to threeHow Tolan layered in in-app purchases and ads for additional monetizationTolan's vision for competing vs. Instinct, Muse, and Grok Bot in Personal AIAjay Mehta:Website: https://www.tolans.com/ LinkedIn: https://www.linkedin.com/in/ajayumehta/ X: https://x.com/ajaymehtaPhil Carter:Website: https://elementalgrowth.com/Substack: https://philgcarter.substack.comLinkedIn: https://linkedin.com/in/philgcarterX: https://x.com/philgcarter

DashPass has stayed at $9.99/month since its launch back in 2018, a choice the person who built it believes is exactly right: Jimmy Liu, the original creator of DashPass.He joined DoorDash in 2016 to run the San Francisco market, back when the company was losing $12 per delivery there, and spent his early months personally emailing every first-time customer to understand what wasn't working. Those conversations became the foundation for DashPass, and Jimmy used them to convince a skeptical leadership team to bet on subscriptions for food delivery years before it became an industry norm.He eventually became DoorDash's Head of Product for New Verticals, and DashPass members now place about 75% of DoorDash's US grocery and retail orders.Before DoorDash, Jimmy founded his own startup called GymFlow, an experience that instilled him with the ownership mentality and hustle that shaped his work at DoorDash from day one.In this episode, Jimmy breaks down:Why DashPass's first version was unprofitable and the three levers that fixed itHow an A/B test that kept compounding proved DashPass was workingHow DashPass helped DoorDash expand into grocery, retail, and other verticalsWhy Jimmy believes keeping DashPass at $9.99/month can maximize long-term growthWhy DoorDash could become a personal assistant for the physical worldHow DoorDash out-executed a better-funded Uber and an earlier-to-market GrubhubJimmy Liu:Website: https://www.doordash.com/LinkedIn: https://www.linkedin.com/in/jimmyliu2/X: https://x.com/jimmayPhil Carter:Website: https://elementalgrowth.com/Substack: https://philgcarter.substack.comLinkedIn: https://linkedin.com/in/philgcarterX: https://x.com/philgcarter

As AI-generated content floods the internet, how do you know what's still human?Weeks after ChatGPT's release, Edward Tian built the first version of GPTZero from a cafe in Toronto over New Year's, tweeted it out, and watched it go viral overnight. What started as his senior thesis project at Princeton became the first commercial AI detector, and three years later has grown into a full authenticity suite, covering AI detection, hallucination detection, and education tools used by millions of students and teachers.Before GPTZero, Edward took a gap year from Princeton to be an investigative journalist who tracked AI-generated disinformation for the BBC. That experience shaped his mission to "preserve what's human on the internet," ultimately inspiring him to start GPTZero.Under his leadership, GPTZero grew profitably from $9M to $30M+ ARR in just 18 months without ever turning off its free tier. The company was recently acquired by Superhuman.On this episode, Edward shares:Why teachers drove GPTZero's early growthThe free "scan box" on GPTZero's homepage and how it drove SEO and GEO trafficHow the company turned viral media moments into a repeatable growth engineHow listicle content helped make GPTZero the top AI detector inside ChatGPT Why he believes editing, not the first draft, is where real critical thinking happens How Superhuman's 40M+ user Chrome extension is expanding GPTZero’s reachThe one thing Edward wishes he had done differently - bought the .com domain earlyEdward Tian:Website: https://gptzero.me/ LinkedIn: https://www.linkedin.com/in/ed-tian/ X: https://x.com/edward_the6 Phil Carter:Website: https://elementalgrowth.com/Substack: https://philgcarter.substack.comLinkedIn: https://linkedin.com/in/philgcarterX: https://x.com/philgcarter

In the crowded subscription app economy, building a great product is no longer enough.As AI has made software significantly faster, easier, and cheaper to build, the primary bottleneck has shifted from product development to distribution.As the serial entrepreneur behind multiple breakout Health & Fitness apps like Fitstar and Zero, Mike Maser has experienced this challenge firsthand. Over the last 15 years, Mike sold his first startup called FitStar to Fitbit, was diagnosed with stage four cancer, made an incredible recovery that he attributes in part to intermittent fasting, and then teamed up with Kevin Rose to build Zero, the world’s #1 fasting app that he has since scaled to $5M+ in ARR and over 10 million registered users. Now, he's navigating Zero through its biggest disruption yet - the rise of GLP-1 medications like Ozempic and Wegovy and the ripple effects they are having on the entire fasting category.Mike breaks down:Why Zero’s simple one-screen MVP was critical for generating early momentumHow a single tweet from Elon Musk drove $1M in revenue for Zero in 24 hoursShifting from organic virality to building a sophisticated growth and monetization engineHow Zero adapted its strategy to the GLP-1 era by helping GLP-1 patients track their hydration and manage their protein intake to maintain muscle massZero’s ongoing strategies to reduce subscriber churn and build lifetime subscriber valueMike Maser:Website: https://zerolongevity.com/LinkedIn: https://www.linkedin.com/in/mikemaser/X: https://x.com/mmaserPhil Carter:Website: https://elementalgrowth.com/Substack: https://philgcarter.substack.comLinkedIn: https://linkedin.com/in/philgcarterX: https://x.com/philgcarter

Every single person alive has asked themselves the same question: "What should I do with my life?"Tomas Hoyos, cofounder and CEO of Apt, is building an AI-powered career mentoring platform to help millions of people answer this question at a moment when AI is about to make it harder for everyone.Tomas breaks down:Why long onboarding quizzes counterintuitively convert better vs. short ones How switching to a conversational AI quiz lifted onboarding conversion 60%+ Why Apt charges from day one but is still betting on freemium long-termThe playbook app-layer AI startups need to win vs. OpenAI and AnthropicTomas Hoyos:Website: https://www.tryapt.ai/Substack: https://tomashoyos.substack.comLinkedIn: https://www.linkedin.com/in/tomas-hoyos/X: https://x.com/tjhoyosPhil Carter:Website: https://elementalgrowth.com/Substack: https://philgcarter.substack.comLinkedIn: https://linkedin.com/in/philgcarterX: https://x.com/philgcarter

When Melissa Cash and her cofounders looked at children's apps, they saw two extremes: addictive games designed to maximize screen time and educational apps built around rote memorization.So they built something different.Pok Pok is an Apple Design Award-winning app offering open-ended, Montessori-inspired play - more like a box of LEGOs than a traditional mobile game. Previously, Melissa designed products for babies and toddlers at Disney and now also works with companies like Solobo Toys and Inkwell Games to create thoughtful play experiences for kids. Tune in to learn: Why Pok Pok deliberately removed levels, rewards, streaks, and every other addictive gaming mechanic from its appHow Melissa thinks about building for kids while selling to parents Why micro-creators are outperforming influencers and traditional paid marketing campaigns for consumer subscription appsWhy the company waited over two years before launching on Android Why the key to building technology for children isn't necessarily less screen time but better screen timeMelissa Cash:Website: https://playpokpok.com/ LinkedIn: https://www.linkedin.com/in/melissa-cash/ X: https://x.com/melissacash_Phil Carter:Website: https://elementalgrowth.com/Substack: https://philgcarter.substack.comLinkedIn: https://linkedin.com/in/philgcarterX: https://x.com/philgcarter

"If you wait until it's perfect, you're too late."This was Tyler Denk’s attitude when he took the growth engine he had built as employee #2 at Morning Brew - the newsletter he helped scale to >3M subscribers and a $75M acquisition - and productized it into a new startup called beehiiv.He launched with a fraction of the features his competitors had, but built an entire process around seeking out criticism, using it to inform the company’s product roadmap, and then executing relentlessly to delight beehiiv’s customers.This ethos, paired with a 0% take rate model that counterpositions beehiiv vs. Substack, has propelled the company to $30M+ in ARR in just a few years, and it's not stopping there.beehiiv’s recent product launches have taken it beyond newsletters to websites, podcasts, and online communities, positioning it to be the operating system for the entire content economy.In this episodeWhy Tyler built beehiiv on a 0% take rate and calls Substack's model "predatory"How building in public has become a secret weapon that propels beehiiv’s growthFeatures in beehiiv’s recent launch, including Community, Copilot, and a new visual editorWhy Tyler believes AI will ultimately make human-created content more valuableTyler DenkWebsite: https://www.beehiiv.com/ LinkedIn: https://www.linkedin.com/in/tyler-denk/ X: https://x.com/denk_tweetsPhil Carter:Website: https://elementalgrowth.com/Substack: https://philgcarter.substack.comLinkedIn: https://linkedin.com/in/philgcarterX: https://x.com/philgcarter

Most AI coaching apps are just ChatGPT wrapped in a personality. Raj Singh and Mark Manson built Purpose to prove there's a better way.Raj, cofounder and CEO of Purpose and previously the founder of GoMoment (acquired by Revinate), explains how they engineered an AI mentor that over 40% of users already call "life-changing" less than a year into the company's existence.In this episodeThe prototype that revealed AI could outperform a $300/hour therapistHow a poker game led to Mark Manson becoming a cofounderThe playbook that took Purpose from launch to $1M ARR in two weeksWhy Purpose initially stayed founder-funded to protect its missionRaj SinghWebsite: https://www.purpose.appLinkedIn: https://www.linkedin.com/in/rajsinghla/X: https://x.com/RajSinghLAPhil Carter:Website: https://elementalgrowth.com/Substack: https://philgcarter.substack.comLinkedIn: https://linkedin.com/in/philgcarterX: https://x.com/philgcarter