Articles by Grant
-
The Design Tax
The Design Tax
A third of Gamma's early team was product designers. That's not typical for an early-stage startup.
35 Comments
Activity
103K followers
-
Grant Lee shared thisJumping on calls with creators late at night. Writing briefs. Failing repeatedly. For months, I was a CEO teaching myself influencer marketing from scratch. Other founders often asked why I didn’t just outsource this or delegate it to someone on our team. I kept going. And what I learned in those months influenced how we built our early team. Every call taught me something. Which creators actually opened the product before our conversation. Which ones could tell our story in their own voice versus reciting a brief. Garry Tan put it well: "You do the stupid manual thing enough times and the real bottleneck just emerges." That's exactly what happened. But the real payoff came later, when I started hiring. After those months doing growth marketing myself, I could interview candidates with real questions. Every question came from a specific way I'd failed. I knew what great looked like because I'd tried and mostly failed to get there myself. If a candidate couldn't uplevel a founder who'd learned the hard way, they weren't clearing the bar. You do the work yourself, so you earn the ability to recognize someone who does it better. What job did you have to learn before you could hire for it?
-
Grant Lee shared thisThere's a hidden tax on every person who has something to say but doesn't have time to produce a video. You know the workflow: write a script, build a deck, find a recording tool, set up the camera, re-record because you said "um" too many times, edit, export, upload. By the time the video is live, half the content is already outdated. Today, HeyGen is eliminating that and they built it on the Gamma API. Here's how it works: you type a prompt or paste a document into HeyGen. Gamma generates the presentation. HeyGen layers an AI avatar on top with auto-generated narration. You get a finished video in minutes. Companies spend weeks turning internal knowledge into polished training videos. Now it's one prompt helping you do work that matters the most. But what excites me most isn't the integration itself. It's what it represents. Gamma started as a product. It's becoming a platform. When another company can take our API and build an entirely new workflow that didn't exist before, that's when you know the platform thesis is real. We're still early. But partnerships like this are exactly why we built the API, and exactly where we're going. Huge congrats to Joshua X., Kevin Raheja, Holly Xiao and the HeyGen team. And to Byron Jones, Deeni Fatiha, and Ravish Agrawal on our side for making this a great co-launch. Try it!
-
Grant Lee shared thisSquare is in a tough chapter right now. But their origin story is one worth revisiting - a lesson in the power of invisible work: It started with a $2,000 glass faucet a woman in Panama wanted to buy with an Amex card. The artist couldn't accept it. She left and never called back. That lost sale eventually became a company Amazon couldn't kill. Jim McKelvey, the artist, called his friend Jack Dorsey. "Why can't small merchants accept credit cards?" So they built Square, a reader that cost 97 cents to make. No contracts. One flat rate. It grew 10% week over week for two years with zero ad spend. Then Amazon launched Amazon Local Register. Undercut Square by 30%. Added live support. Distributed through Staples. McKelvey knew his costs to three decimal points. His first reaction: "That math doesn't work." Amazon had priced against Square's rate without realizing it was already razor thin. They were underwater on every transaction. Square did nothing. Amazon had copied the visible product. But over five years, Square had built what McKelvey describes in his book The Innovation Stack: 14 interdependent innovations - pricing, fraud models, hardware, settlement speed - each solving a problem created by the last. None planned. All compounding. Amazon was looking at a finished product. They couldn't see the system underneath it. Their version lasted 14 months. When Amazon killed it, they needed somewhere to send their stranded merchants. They partnered with Square to mail them free Square readers. In smiling cardboard boxes. Square went public three weeks later. You don't plan an innovation stack. You earn it. What would a competitor need to live through to copy what you've built? Answer that question, and maybe you've found your moat.
-
Grant Lee shared thisMy son made a hat at a birthday party. Loved it. Showed it to everyone. A week later, it was gone. Never mentioned it again. But his Messi jersey - he wore it four days a week for an entire year. He didn't make it, but it means the world to him. Both made him happy. They just served different needs. If you're running a SaaS company right now, you're hearing the noise. "AI will kill SaaS." "Anyone can build software now." "The subscription economy is over." Vibe coding platforms are generating tens of millions of projects. Some are weekend experiments. Some are experienced engineers shipping in hours what used to take weeks. The tools are real, the output is accelerating, and it's becoming its own massive category. But it's a different category from the software your customers reach for every day. Harvard researchers called it "The IKEA Effect." People who build something themselves value it 63% more than a pre-built version. That's a real force, and it explains why DIY software will keep growing - for hobbyists and professional builders alike. It also highlights why the tools people didn't build but still reach for daily earned that place through something else entirely: - The login that remembered their context. - The update that didn't break their workflow. - The edge case you handled before they knew it existed. Every one of those is a decision your team made well, compounding over years. Speed of creation is a different advantage than depth of accumulated trust - and your customers feel the difference every time they open your product without thinking twice. The SaaS market is projected to nearly double by 2030. DIY software will grow into its own massive category. Both things are true, serving different needs. Hats are fun to make. Jerseys are hard to replace. Just keep building.
-
Grant Lee shared thisEvery day, people leave your app because they don’t get it. And they often blame themselves. While building Slack, Stewart Butterfield kept a $14 paperback on his desk called “Don't Make Me Think.” On Lenny's Podcast, he explained why: Think about a time you helped your parents use an app. They tap something. It doesn't do what they expected. The first thing they say isn't "this app is broken." It's "what am I doing wrong?" The slight lean away from the screen. The tentative slower tapping. The quiet apology in their voice. They don't file a ticket. They just decide: this isn't for me. Your users feel this too. They just leave quietly. Larry Tesler, who invented copy-and-paste at Xerox PARC, called it the "Law of Conservation of Complexity": every app has irreducible complexity, and the only question is who deals with it. You land, and every step feels like a step you meant to take. Be opinionated on behalf of your user. Absorb the complexity so they don't have to. Where in your product are users quietly blaming themselves?
-
Grant Lee shared thisWe won Product of the Day, Product of the Week, and Product of the Month on Product Hunt. Signups were pouring in. Yet we could feel it wasn't working. No compounding. No momentum. We sat down as a founding team and forced ourselves to be honest. "This isn't going to be enough." Scott Belsky (founder of Behance, former CPO of Adobe) has a rule I think about constantly: in those first seconds, users are lazy, vain, and selfish. Design for that instead of fighting it. We rebuilt with this in mind: one moment of magic, delivered in seconds. Signups went from a few hundred a day to thousands. Then 5,000. Then 10,000. Then 20,000. No marketing. No ads. Just a product that earned its next 30 seconds. What would change if you asked that same question about your product - "Have we earned the next 30 seconds?"
-
Grant Lee shared thisYour ideas are being dismissed for the most avoidable reason. If two people walk into a meeting with equally strong insights, the one with better designed materials is taken more seriously. That’s not a matter of opinion, it’s a well-documented cognitive bias. This is a distortion in almost every meeting, pitch, and presentation where ideas are not judged on their quality alone, but on how well they are packaged. And yet, most of the people being judged in this way were never trained to package ideas visually in the first place. I call this The Design Tax. A study of over 1,000 office workers found that the average employee spends 20 hours a month on presentations, with more than 40% of that time going to formatting alone. That’s an entire workday, every month, spent on something that rarely makes the actual ideas better. I’ve experienced this firsthand. Late nights spent adjusting layouts, aligning text, and reworking slides, just to make sure the thinking wasn’t dismissed because of how it looked. It’s a disproportionate amount of effort that takes time away from the work itself. The design tax is paid by everyone. People who never claimed to be designers, but are judged as if they are. This is one of the most empowering elements of AI. AI has the capacity to remove the expectation that everyone else needs to operate as a designer just to communicate clearly. I wrote more about this in the latest newsletter, including why I think the companies that understand this shift will have a structural advantage. Gamma is positioned to cut this tax once and for all. You can read my full piece below.
-
Grant Lee shared thisIn the 1980s, manufacturers automated away apprenticeship programs to cut costs. Forty years later, 1.9 million manufacturing jobs sit unfilled and the economy bleeds a trillion dollars a year. White collar companies are making the same mistake right now. Entry-level job postings have fallen 35% since January 2023. New grad hiring at Big Tech is half what it was pre-pandemic. McKinsey is deploying 12,000 AI agents to handle work once done by junior consultants. The logic feels obvious. AI can build the report, draft the memo, summarize the research. Why pay someone to do what a tool does in seconds? Because the report was never the point. Jensen Huang washed dishes on the graveyard shift at Denny's starting at age 15. Cleaned toilets. Bused tables. He built a personal efficiency system: "I never left the station empty-handed. I never came back empty-handed." Today he runs NVIDIA with 50+ direct reports and a culture where no task is beneath anyone. You can draw a straight line from the dish pit to the org chart. Junior white-collar employees are on that same line. Every deck built, every dataset cleaned, every memo drafted is a rep. Individually, low-value. Collectively, they build pattern recognition and judgment that no shortcut can replicate. Cisco President Jeetu Patel called refusing to hire entry-level workers because of AI "the stupidest thing a company can do." Premium wages can't fix a pipeline that was never built. The consequences won't arrive this quarter. They'll arrive when companies discover a hollowed-out middle layer and scramble to hire experienced people who don't exist, because nobody invested in making them. I understand the pressure. Margins are tight. AI is genuinely capable. Investing in junior talent right now feels like catching a falling knife. But the companies that catch it will have something their competitors won't: leaders who grew up inside the business, who built judgment the only way it gets built. Every expert was once a beginner someone believed in. Who's leading your company in 2036 if you stop investing in them now?
-
Grant Lee shared thisJeff Bezos said something at a shareholder meeting back in 2011 that startups should consider: "If you invent frequently and are willing to fail, then you never get to that point where you really need to bet the whole company." One sentence that explains more company failures than most business books. A Kodak engineer invented the digital camera in 1975. Management's response: don't tell anyone about it. They held over 1,000 patents in digital imaging, commanded 90% of the US film market, and still filed for bankruptcy. Kodak didn't make too many mistakes. Maybe they made too few. They optimized for protecting small, steady wins and avoided the messy, error-filled work of reinvention. Until they reached exactly the point Bezos warned about; needing to bet the whole company. This mindset shapes how we try to build at Gamma. Break every big bet into small, testable pieces. None of the ideas we test early are massive. A bunch die right away. The ones that pass, you build bigger. By the time you're shipping, you've gone through layers of testing and iteration. Compress timelines to force the learning. You're going to make mistakes; you just don't know which decisions are the mistakes yet. Speed is how you find out. Your mistake rate is noise. Your learning rate is signal. What would you build if small failures didn't scare you?
Grant Lee commented on a post
3d
Congrats and big thanks to the HeyGen team. Really like how both products complement each other here. They're each doing what they do best and it just works.
Grant Lee commented on a post
3d
This is super cool, and it has been fun to watch our teams partner on this. A completely new workflow that didn't exist before this integration.
Grant Lee commented on a post
3d
Thank you for the amazing partnership Joshua X.! It's been so fun working with you and the team on this.
-
Grant Lee liked thisGrant Lee liked thisBuilding a product is one thing. Branding it is a completely different act of love. The product answers “does it work?” The brand answers “do I care?”. One of my favorite things right now is crafting the visual identity for products I’m shipping. Not outsourcing it. Not templating it. Sitting with the essence of the thing and designing how it shows up in the world. This is Pali — a generative platform for thinkers to explore, merge, and ideate on concepts before handling the execution to agents. The product is alive. Now it needs a soul :) Still cooking all the features behind the scene :)
-
Grant Lee reacted on thisGrant Lee reacted on thisLast Monday, Alice Walton and Heather Murray were invited to join an event hosted by Gamma... ...keeping the rest of us impatiently waiting for updates and photos. Gamma hosted an early-morning fireside chat at SecondHome with Clay and Granola, full of brilliant conversations about how AI is changing the way businesses work. We were excited to hear some great takeaways: 👉 "Superpower energy." Diana Kimball Berlin's phrase for what AI does - it amplifies what people can do. A statement close to our hearts. 👉 Show up with proof. Clay's "reverse demo" idea - arriving with custom data and proof of work before anyone's even asked is such a smart way to cut through a noisy market. 👉 Do one thing brilliantly. Granola's whole approach is to solve one problem really well and resist the urge to add more. No feature creep, just focus. (We could all do with a reminder of that one.) Alice and Heather also got to chat with Grant Lee afterwards and came back buzzing about what Gamma are building long-term and their approach towards keeping the entire team engaged in yearly hackathons. It was great to hear that he was impressed with our work and enthusiastic about the AI for Non-Techies mission to make AI accessible for everyone. What a lush Monday! What's the best event you've been to this year?
-
Grant Lee reacted on thisGrant Lee reacted on thisMore chronicles of my journey at Strathens while I’m currently in Belgium. Launched our beta more than 2 weeks ago, hundreds of users but we had no onboarding. Users were entering, but without knowing what to do next. We made the error of thinking that the user will perceive / understand the platform the same way we did. But definitely not. We quickly realised we had to change that and make the first 30 seconds the best-in-class moment for a user. Make them feel they actually found the exact solution to their pain. We based our onboarding on the principle described by Grant Lee from GAMMA APP, definitely one of the best onboardings I saw in a self-service AI platform — https://lnkd.in/eqg2AFBt Founders and creators, don’t forget that your product it’s not a priority for users!
-
Grant Lee liked this🫡Grant Lee liked thisI see a lot of founders making the same mistake. They hire the wrong branding partner. Usually a big agency that takes way too long. The best brands I’ve seen don’t come from big agencies. They come from small, sharp teams with super strong taste. Teams that move fast and build with you, not for you. So I curated my go-to list of studios I’d recommend to any early-stage founder building in consumer tech. Built in Lovable. https://lnkd.in/dUYTSmGx
-
Grant Lee reacted on thisGrant Lee reacted on thisIf you want to rest intentionally, you need to release the guilt. Resting is not stealing time it’s restoring it. But for a lot of us, rest doesn’t feel natural. It feels uncomfortable. Because somewhere along the way, we were taught that our value is tied to what we produce. It starts early. At school, we’re rewarded for performance. Good grades, praise, being “on top of things.” We learn quickly: doing more = being valued. Then we grow up, and ambition gets framed as long hours, side projects, and always being switched on. Being hard-working becomes a badge of honour. Slowing down? That starts to look like laziness. So when we try to rest, it’s not just about stopping. It becomes a question of worth: “If I stop… am I still enough?” That’s why rest isn’t just about time it’s about trust. Trust that things won’t fall apart if you pause. Trust that your value doesn’t disappear when you’re not producing. Trust that you don’t have to constantly prove yourself. This is something I’ve had to unlearn and I’m still unlearning it. Which is exactly why I created my Rest vs Ambition Planner. Not to help you do more. But to help you hold both. To plan your ambition and prioritise your rest without guilt. Because you don’t earn rest. You need it. If you’re trying to build a life where rest and ambition can coexist, this is for you. Link in bio 👇
-
Grant Lee reacted on thisGrant Lee reacted on thisAI can improve individual tasks. But when does it actually improve firm performance? We study a key bottleneck in AI adoption inside firms: what we call the "mapping problem" -- discovering where and how AI creates value within a firm's production process. In a field experiment with 515 high-growth startups, we find that helping firms solve this "mapping problem" led them to discover 44% more AI use cases, complete 12% more tasks, become 18% more likely to acquire paying customers, and generate 1.9x higher revenue -- with the biggest gains in the upper tail. AI didn't only marginally improve the average firm. It expanded what the most promising ventures could achieve. At the same time, treated firms demanded 40% less external capital and did not increase labor demand proportionally -- suggesting they can do more with less. For researchers, the paper speaks to a central question: when do AI's task-level gains aggregate to firm performance? We show that AI can improve firm performance even at its current capabilities, and that there may be an earlier friction before investment costs and lags: discovering how to reorganize. For founders, managers, and policymakers, one simple takeaway is that access to the tools is not enough; firms need to learn how to redesign their production process around AI. I created a Gamma to summarize the findings (cc Grant Lee :)), and more details are in our working paper here: https://lnkd.in/g8xbwSfb This project was a joy to work on with my amazing coauthors Dahyeon Kim and Rem Koning. And a big thank you to the many people at INSEAD who made this project possible!
Experience & Education
-
Gamma
View Grant’s full experience
See their title, tenure and more.
or
Languages
-
English
Native or bilingual proficiency
-
Chinese
Elementary proficiency
Recommendations received
4 people have recommended Grant
Join now to viewOther similar profiles
Explore more posts
Explore top content on LinkedIn
Find curated posts and insights for relevant topics all in one place.
View top contentOthers named Grant Lee
306 others named Grant Lee are on LinkedIn
See others named Grant Lee