Annotations (28)
“Because we didn't come from performance marketing or Facebook, we had no preconceived notions of how things should go. And when you talk to Facebook marketers who are moving to TikTok, they were applying all these rules and heuristics of Facebook and Instagram to TikTok. But these are two separate platforms. You have this Chinese app and a different algorithm. Why would those rules matter over here? I spoke at an all-hands for TikTok.”— Greg Stewart
Solving TikTok · p. 15
Strategy & Decision Making · Psychology & Behavior
DUR_ENDURING
First principles beats imported heuristics
“We spent a lot of time, the first two, three months I was there trying to dissect the current business. And the current business was this one-to-one marketplace connecting an independent contractor coach with a consumer who was looking for personal training. We're studying the behavior and watching what's going on here. And what I noticed was that coaches could set their own price. Some people have crazy high prices. Some people have more approachable prices.”— Greg Stewart
How One Coach Changed Everything · p. 8
Business & Entrepreneurship · Psychology & Behavior · Operations & Execution
DUR_ENDURING
Fake personalization: big bucket personas
“Most companies in our space, they're started by creators. And they're good products. They're good companies. But the creator is the face of that business. They make every decision, and they don't have a DNA that's rooted in engineering and problem solving. And when you look at these apps, they're mostly just content libraries. And the motion is just constantly creating more and more content.”— Greg Stewart
The Duolingo of Fitness · p. 2
Strategy & Decision Making · Business & Entrepreneurship · Technology & Engineering
DUR_ENDURING
Engineering DNA beats content treadmill
“We looked at apps like social networks and we looked at Duolingo, all these apps that we're using, all these powerful motivational mechanics and pointing them at an action. For social media networks, it's selling your attention. For Duolingo, it's learning a language. And so we took that mentality of how do we use software to create an experience that's totally different than what exists today, that isn't reliant on a never-ending content machine.”— Greg Stewart
The Duolingo of Fitness · p. 2
Creativity & Innovation · Psychology & Behavior · Technology & Engineering
DUR_ENDURING
Borrow engagement mechanics across categories
“Lauren went out to her Instagram audience and we had like a 100 people sign up at a 100 bucks a month, very fast. That was the most exciting thing I'd seen yet. And so that first month, we were just watching what was happening. The most amazing and impressionable thing to me was these people had found each other in the app. They were similar personas. They lived in a similar place, busy women following the same workouts every day, interacting with the coach and each other.”— Greg Stewart
How One Coach Changed Everything · p. 9
Business & Entrepreneurship · Psychology & Behavior
DUR_ENDURING
Organic community: strangers meeting in parks
“You have to make a clear, clear case of what this does to the business. We have no Android app. People think that's insane. But building an Android app requires basically pausing development on iOS, splitting the team's focus in time, playing this game of catch-up for a user that has much lower revenue potential and could absolutely take the business sideways for a year. Will we have an Android app? Yes. We won't have it this year. It's these levels.”— Greg Stewart
Ladder as a Single System of Record · p. 22
Strategy & Decision Making · Operations & Execution
DUR_ENDURING
No Android app: ruthless prioritization
“We built a product called Ladder Pulse that automatically comes in, reads every single chat that's come in, tells you, here are the three most burning questions. Here's the content you should create. Here's the member you should respond to because they've never been responded to by a coach. Send. Done. So now we've removed all the cognitive overload of what to say. And it's all powered by this tool that we built purposely for the business.”— Greg Stewart
AI & Other Potential Growth Opportunities · p. 21
Technology & Engineering · Operations & Execution
DUR_ENDURING
AI removes coach cognitive overload, preserves human touch
“The nutrition survey, that was probably a couple 100 questions. We read all of them. Now you can put them in ChatGPT. In the beginning of this, I would copy and paste App Store reviews into a Word doc. I would read every single one of them. And I'd organize the words by buckets. I would color code them. And I would do ChatGPT. And I have these documents that are a 100 pages long. And it's all just deconstructed words from our members. And very similar to survey work.”— Greg Stewart
Don't Guess, Ask · p. 11
Operations & Execution · Business & Entrepreneurship
DUR_ENDURING
100-page customer feedback synthesis docs
“Programming, you know exactly what to do. Someone's prepared it for you. There's no guesswork. There's no thinking. Coaching, you have an expert there to guide you, answer your questions. And accountability, if a coach is standing in front of you, and you don't want to piss off your coach, he's a really powerful motivator. Then we took those three pillars and designed an experience from the ground up to get as close as possible to that experience.”— Greg Stewart
The Duolingo of Fitness · p. 1
Business & Entrepreneurship · Psychology & Behavior · Operations & Execution
DUR_ENDURING
Programming, coaching, accountability framework
“What people don't realize in consumer is how freaking hard it is. And there is no quick fixes. There is no growth hacks are not a real thing. You have to be black belt at building products for the consumer and growth. You have to have both those things within your own team to survive and then be able to raise money to go fund that, which is its own mission. Half our team worked on workout completion. Half of them works on trials off TikTok. That's the business. Very simple.”— Greg Stewart
Why Building Consumer Businesses is Fun · p. 19
Business & Entrepreneurship · Strategy & Decision Making
DUR_ENDURING
Consumer requires black belt in both product and growth
“We started handles in the coach's names. Then we started to create content. And very, very quickly, we started to learn what was the content that worked. And we were dissecting every inch of it. What worked multiple times in a row? Why? What are the commonalities? How is this product used? It's not Instagram. It's like TV. People are consuming content for entertainment. And it's not a social platform. It's a media company.”— Greg Stewart
Solving TikTok · p. 14
Business & Entrepreneurship · Creativity & Innovation
DUR_CONTEXTUAL
TikTok is media company, not social
“What we learned about hooks, what you say in the beginning, the first three sentences. It matters. What should a hook look like? What do I need to say? It gets a billboard that I need to get your attention very quickly. And we had a lot of bad content that didn't work. And then we had stuff that started to work. And we have these whiteboards where we dissect every inch. It's like, what is she wearing? What word was first? What was the setting of the gym? What was the movement? What did we say?”— Greg Stewart
Creative Control & Building Awareness · p. 16
Business & Entrepreneurship · Creativity & Innovation · Psychology & Behavior
DUR_ENDURING
Whiteboard dissection: every variable matters
“We built a customer support tool from scratch. That person's name is Maeve. We built Maive AI. And we built a customer support tool from scratch. It's purpose-built for our company that manages now 90% of the flow coming in. And the experience is as good or better. And it's faster than it was before. And we made it ourselves in our team. That would not have been possible five years ago. That solved the huge problem.”— Greg Stewart
AI & Other Potential Growth Opportunities · p. 21
Technology & Engineering · Operations & Execution
DUR_CONTEXTUAL
Custom AI support handles 90% of tickets
“One, don't listen to any one person on product. Everyone has a unique use case, persona, problems, solutions that they're looking for. So no one person is the source of truth. If you get one piece of feedback, that's just one piece of feedback. So when we talk about investors, I remember earlier meetings, it would be like really prescriptive advice saying, you should go build this. And it's like, maybe. I don't know. That's what you want.”— Greg Stewart
Don't Guess, Ask · p. 12
Strategy & Decision Making · Business & Entrepreneurship
DUR_ENDURING
Don't build what one investor wants
“The only way to get this going is to write the first check and lead with that conviction, showing that you had skin in the game here again. At this point, I'm trying to figure out how to sell anything that's not nailed down. And so it's 401k, put that in, done. What else can we do?”— Tom Digan
Fundraising 101: Selling Conviction · p. 7
Business & Entrepreneurship · Psychology & Behavior
DUR_ENDURING
Liquidate personal assets to lead round
“Don't guess. We absorb every piece of information that's coming in and it's all driven by our members. And if you ask the right questions and you have people that are talking to you, you can start forming a picture of what are the huge buckets that actually move the needle. And you can do some work to figure out how complex is this to go build and go do it. And don't do 10 other things just because it's interesting or you can do it. Just do the one thing and do it really well and do it again.”— Greg Stewart
Don't Guess, Ask · p. 10
Strategy & Decision Making · Operations & Execution
DUR_ENDURING
Don't guess; ask customers ruthlessly
“One of the biggest lessons, particularly from that moment, is selling people on your conviction. In that moment, we were selling both the product that we were about to shut down, but we didn't really even have visuals yet for the new product. So it was as much just selling the team and the conviction on what you were about to build.”— Greg Stewart
Fundraising 101: Selling Conviction · p. 7
Business & Entrepreneurship · Psychology & Behavior
DUR_ENDURING
Sell conviction when product doesn't exist
“In the beginning, it was a non-coding tool. It was how do I synthesize information? Well, it was amazing. I could take 5,000 responses and really understand what's happening. And so then it was everywhere. I can take it to think about how do I create compelling TikTok hooks based on all of this survey data or user data. That's a tool. It saves time.”— Greg Stewart
AI & Other Potential Growth Opportunities · p. 20
Technology & Engineering · Operations & Execution
DUR_CONTEXTUAL
AI prevented team bloat at scaling moment
“I was thinking about what a leadership transition could look like. I found myself in a position where I was our largest financial shareholder. In addition to being in the company, as a co-founder, early days, it was all people like yourself that had put money in because they trusted me personally. So I felt naturally this obligation and my role and mindset shifted from one of co-founder to trying to be a steward of the business.”— Tom Digan
Ladder's Early Chapters · p. 4
Leadership & Management · Psychology & Behavior
DUR_ENDURING
Founder to steward identity shift
“He goes all mad scientist on trying to just understand every aspect of TikTok to the point where, fast forward two years later, and the big dog engineers from China are in our office. Because Greg would be calling and being like, the algorithm is broken. And they'd be like, no, it's not. And then Greg would go get a bunch of information from Ryan Mott and send it through. And they'd be like, oh, wait, he's right. There's something wrong with ads manager because the ad tech was very new.”— Tom Digan
The Cave Process · p. 14
Operations & Execution · Technology & Engineering
DUR_ENDURING
Mastery so deep he found TikTok bugs
Frameworks (3)
Personal Training Value Decomposition
Breaking high-touch services into automatable components
Personal training delivers value through three distinct pillars: programming (what to do), coaching (how to do it), and accountability (motivation to do it). By decomposing a premium service into its component parts, software can replicate most of the value at scale. The key insight is that what appears to be a single service is actually a bundle of separate functions, some of which can be automated while preserving most of the customer value.
Components
- Identify the high-touch service to decompose
- Break the service into component functions
- Determine which components can be automated
- Build the automated components first
- Use technology to simulate the human components
Prerequisites
- Deep understanding of the premium service being disrupted
- Engineering capability to build software
- Access to experts in the domain
Success Indicators
- Customers achieve 70%+ of the results they'd get from the premium service
- Unit economics are 10x better than the premium service
- Retention rates are comparable to or better than premium service
Failure Modes
- Automating the wrong components first
- Removing so much human touch that customers feel the product is soulless
- Failing to validate that customers will accept the automation
Customer-Driven Product Development Loop
Building products by absorbing customer signal, not guessing
Instead of guessing what customers want or copying competitors, systematically absorb every piece of customer information available. Organize that information into high-impact buckets. Assess build complexity for each bucket. Then ruthlessly focus on building one thing at a time, the thing with the highest impact-to-complexity ratio. Repeat. This framework eliminates guesswork and ensures you're always building the right thing next.
Components
- Absorb all customer information
- Identify the high-impact buckets
- Assess build complexity for each bucket
- Ruthlessly focus on one thing
- Ship, measure, iterate, repeat
Prerequisites
- Access to raw customer feedback channels
- Willingness to read hundreds or thousands of responses
- Discipline to focus on one thing at a time
Success Indicators
- Product roadmap is driven by customer data, not founder intuition
- High feature adoption rates (customers actually use what you ship)
- Increasing retention and satisfaction scores
Failure Modes
- Founder doesn't personally read customer feedback
- Team builds multiple things simultaneously instead of focusing
- Analysis paralysis: too much absorbing, not enough building
Three Pillars of Sales
Likability, trust, and persistence as the foundation of selling
Effective sales is not about tricks or tactics. It's about three foundational elements: making people like you (ideally to the point where they root for you), building trust through integrity (because that's all you have), and being deeply persistent without being annoying. These three pillars work together: likability gets you in the door, trust gets you the deal, and persistence gets you through the inevitable objections and delays.
Components
- Build likability
- Establish trust through integrity
- Be relentlessly persistent
Prerequisites
- Genuine interest in people
- Commitment to integrity
- Willingness to play the long game
Success Indicators
- People return your calls
- Deals close because of the relationship, not just the product
- Customers become advocates and refer others
Failure Modes
- Faking likability instead of being genuine
- Cutting corners on integrity to close deals faster
- Being annoying instead of persistent
Mental Models (3)
Counter-Positioning
Strategic ThinkingTaking a strategic position that incumbents cannot or will not copy because it would undermine their
In Practice: Greg Stewart explaining how Ladder's engineering-first approach differed from creator-led competitor
Demonstrated by Leg-td-001
First Principles Thinking
Decision MakingReasoning from fundamental truths rather than by analogy or convention. When Facebook marketers told Greg to wait two weeks during the TikTok 'learning phase,' he ignored them and changed budgets 7-10 times per day based on observed results. He reasoned from first principles: if something works, do more of it immediately. If it doesn't work, stop immediately. Don't follow rules imported from a different platform.
In Practice: Greg Stewart explaining how he ignored Facebook-derived rules when optimizing TikTok ads
Demonstrated by Leg-td-001
Opportunity Cost
EconomicsThe cost of any decision is what you give up by choosing it. When Ladder chose not to build an Android app, the opportunity cost was potential Android users. But the benefit was maintaining iOS development velocity and focus. The decision to defer Android was explicit recognition that building it would cost more (in lost iOS momentum) than the immediate revenue gain was worth.
In Practice: Greg Stewart explaining the decision not to build Android app despite pressure
Demonstrated by Leg-td-001
Connective Tissue (3)
Duolingo's motivational mechanics applied to fitness
Ladder borrowed the engagement mechanics that make Duolingo addictive (streaks, badges, celebrations, progress tracking) and pointed them at strength training instead of language learning. The insight is that powerful motivational systems from successful apps in other categories can be transplanted to new domains. Duolingo gamified language learning; Ladder gamified fitness. Both use software to create behavior change through systematic positive reinforcement and social proof. The parallel shows that engagement mechanics are domain-agnostic and transferable across categories.
Greg Stewart explaining Ladder's product strategy and how they looked to apps like Duolingo for inspiration on engagement mechanics
TikTok as media company vs. social network architectural difference
TikTok's architecture is fundamentally different from Instagram or Facebook. Social networks are organized around a social graph where you follow people and see their content. TikTok is organized around an algorithmic content graph where the platform decides what you see based on engagement patterns. This makes it more like a TV network than a social network. The implication for creators and brands: on Instagram, you create content for your audience. On TikTok, you create content that the algorithm will recognize and distribute to the right audience. The architectural difference determines entirely different content strategies. Ladder succeeded on TikTok because they understood this fundamental difference while competitors were treating it like Instagram.
Greg Stewart explaining how TikTok's mechanics differ from social networks and why that mattered for their content strategy
Trading as mental model for performance marketing
Greg Stewart compared managing TikTok ad budgets to trading. Both require rapid decision-making based on real-time data, both have screens with multiple metrics updating constantly, both reward quick pattern recognition and willingness to cut losses fast. A trader moves capital between positions based on performance signals; a performance marketer moves ad budget between campaigns based on performance signals. The parallel suggests that trading skills (speed, detachment, data-driven decisions, risk management) translate directly to performance marketing. The cognitive load, the decision velocity, and the need for emotional discipline are identical.
Greg Stewart describing his experience managing TikTok ad spend and comparing it to trading
Key Figures (4)
Greg Stewart
45 mentionsCEO
Tom Digan
12 mentionsCo-founder and President
Tom Digan co-founded Ladder after leaving a successful hedge fund career.
- Startups are really hard and it was messy. We were kind of out of money the whole time.
Lauren (Ladder Coach)
8 mentionsPersonal Trainer and Creator
Edsel (Creative Director)
5 mentionsHead of Branding and Creative
Glossary (2)
creditors
DOMAIN_JARGONEntities owed money by a debtor
“We were in debt and owed a lot of people a lot of money.”
CAC
DOMAIN_JARGONCustomer Acquisition Cost; the cost to acquire one customer
“The capital markets aren't funding CAC in consumer companies.”
Key People (2)
Brent Beshore
Permanent equity investor who buys and holds family businesses
Ken Griffin
(1968–)Founder and CEO of Citadel
Concepts (3)
ARR (Annual Recurring Revenue)
CL_FINANCIALPredictable annual revenue from subscriptions or contracts; key SaaS metric
Product Market Fit
CL_STRATEGYThe degree to which a product satisfies strong market demand; when customers pull the product
Learning Phase (Ad Platform)
CL_TECHNICALPeriod when ad algorithms gather performance data
Synthesis
Synthesis
Migrated from Scholia