Luis von Ahn on scaling Duolingo and advising early-stage founders

Executive overview

Three early-stage founders get live coaching from Duolingo CEO Luis von Ahn on the hardest early questions: when to scale, which product to back, and how to price. The recurring pattern across all three: founders undervalue their work and delay monetisation longer than they should.

Charging more — not raising capital — is usually the first lever founders should pull.

Going public changed how Duolingo operates

• Private company = no outside scrutiny; public company = forced operational discipline
• Forecasting became precise and constrained hiring decisions productively
• Stock performance validated the decision, but the bigger gain was internal rigour

Watch behaviour, not what people say they want

• Users said they wanted live human tutors; in practice, almost nobody uses them
• Survey answers reflect aspirations; product decisions should track actual behaviour
• Duolingo competes for time against TikTok and Instagram, not other language apps

Case 1: Good Things Vending — art vending machines in Chicago

• Refurbished machines stocked with local art, stickers, and pins; $20 and under per item
• 7–8 machines, $100k gross revenue, 53% YoY growth in Q1 2025
• Question: when to scale fast vs. build sustainably?
• Excess demand is a position of strength — use it to add friction (waitlists, placement fees)
• A placement fee weeds out low-commitment venues; it is a signal of care, not just revenue
• Opening a second city clarifies what is truly core to the business vs. what is just Chicago
• Next hire should own operations so the founder can focus on expansion strategy
• Franchising reluctance is valid; focus on making the Chicago model excellent — good things get copied

Case 2: Behr — premium Apple accessories

• Cable organiser (Tame) and Vision Pro bag (Tote) made from merino wool, leather, machined metal
• Tote stalled because Apple Vision Pro adoption stalled; Tame left as unintended hero product
• Question: how to validate real demand before over-investing?
• Do not build for unproven platforms — Apple Watch apps failed despite Apple's promises
• Be present on established platforms; being early is not an advantage in hardware accessories
• Saturated markets are not disqualifying; differentiated materials (merino wool) can stand out
• Seed product with local creatives and influencers first — Instagram launched with photographers only
• Energy is better spent on product-market fit than secondary brand signals like scent packaging

Case 3: NatureSpiral Landscapes — biophilic interior design

• Living walls, moss walls, interior plantscapes for offices, apartments, public spaces
• Five staff plus subcontractors; approaching $1M revenue; thin margins despite strong demand
• Question: how to fund headcount when revenue is high but profit is low?
• Raise prices first — design and expertise are not easily replicated; 50%+ margins are achievable
• Higher prices can increase perceived desirability in premium design categories
• Inconsistent cash flow is the core problem; pricing fix addresses root cause better than a loan
• An SBA loan is a reasonable tool given established track record and assets — not a last resort
• Additional revenue streams (plant shop, cafe, space rental) are worth exploring once pricing is corrected

Luis von Ahn's advice to his 2012 self

• Monetise earlier — Duolingo waited until 2017; revenue enabled mission, not just profit
• Accept your own weaknesses instead of fighting them; hire to cover gaps in your executive team
• Most of Luis's leadership team excels at things he is genuinely poor at — that is intentional