digest
Short Summary
Main recommendation: Build a minimum evolvable product, not just a minimum viable one, and find your first users through targeted search for people with a burning problem or a taste for early adoption, rather than broad persuasion. The video argues that early adopters are rarely price sensitive, that paying customers give sharper feedback, and that early users steer how the product evolves over time.
Key Points
- Early adopters are rare: most people have used zero products from among a startup's first 10 users, so finding them is a search problem more than a persuasion problem.
- Charge real money early, since paying customers give sharper feedback and their anger or praise is more useful than a free user's silence.
- Reach early users through personal outreach such as targeted cold email or direct contact, because billboards rarely reach them.
- Launch early and study early users like an anthropologist to understand how they decide and why they trust you.
- Run constant experiments on pricing, landing pages, onboarding, and features, and accept churn, since startups can fail quietly without public headlines.
Details Worth Noting
- Concrete instructions:
- Search for people who either love trying new products (the "Gustaf" type) or have a burning problem your product solves (the "Ankit" type).
- Charge real money from the start, even though the goal is feedback rather than revenue.
- Use targeted cold emails or direct outreach rather than mass advertising.
- Launch early to create a wide surface area for early users to discover you, even before you know much about them.
- Study early users' decision-making and what they want, treating them like a hidden civilization.
- Run frequent experiments on pricing, landing pages, onboarding, and features, and talk to early users to make them love the product.
- Don't worry excessively if you lose or annoy a user; personal relationships are usually repairable, and there are many others yet to hear of you.
- Build the first version as a minimum evolvable product: simple enough to be exposed to market pressure, flexible enough to adapt based on what early users push it toward.
- The speaker's own example: when his team needed a first inference API without dealing with billing or public endpoints, he found and paid a startup within three days and became its first customer, judging the vendor by whether it solved his problem rather than by its size or reputation.
- Consumer apps can struggle because ads often don't cover AI costs and personal software budgets are small (the speaker estimates his own personal software spend at about $150 a month), so many AI founders start by selling to prosumers, businesses, or high-value users such as doctors.
- Tesla's Roadster is presented as an early-adopter search: the speaker argues it was a way to find buyers willing to pay $150,000 for an impractical car, and that early adopters' preference for tech and acceleration over comfort shaped the Model Y's characteristics.
Bottom Line
Your first users are less about convincing the crowd and more about finding the few people with a real need or appetite for novelty, and their feedback will shape your product's direction, so start simple and let them guide it.
