newsfilter.io
Fireside Chat, Interview

When to Launch Your Startup and When to Wait

  • Core Thesis: YC partners Harj and Brad identify that founders frequently delay launches by over-optimizing for a "perfect" launch event, whereas the successful strategy is to launch early, move quickly, and iterate based on real user feedback rather than internal simulations.
  • Psychological Barriers to Speed

    • Misaligned Velocity Perception: Founders who previously worked at large corporations often mistake "corporate speed" for startup speed, failing to recognize that "startup fast" requires a higher degree of discomfort and improvisation.
    • The "Oscar Ceremony" Fallacy: Founders visualize launches as high-stakes, singular celebrity events where they must look polished; in reality, most launches have no audience, and founders often launch the same product multiple times before gaining traction.
    • Self-Deception: Founders convince themselves they are an exception to the rule ("I know the advice is good, but my case is different"), leading them to optimize for a hypothetical "parallel universe" of perfect reactions rather than reality.
    • Specific Fears: Common anxieties driving delays include fear that users will notice product ugliness, competitors will discover the idea, or investors will see an unfinished product.
  • Case Studies: Hacked vs. Polished

    • Instacart Example: Early in YC, Instacart demonstrated a product with no backend infrastructure; orders were manually fulfilled by founders or friends driving to stores, yet the user interface worked effectively.
    • Brexit Example: Conversely, Brexit had a fully functional virtual credit card system at its core but initially lacked any user interface to display spending history or transaction details.
    • Magic Company: Launched two days after ideation with a simple website and phone number instructing users to text requests, demonstrating the value of learning live rather than waiting for a "ready" product.
  • Addressing Counter-Arguments (The "Exception" Trap)

    • Rippling (Parker Conrad): Often cited as an exception for taking 18 months to build before launch; however, this was only viable due to Conrad's deep domain expertise and massive user base from his previous venture, Zenefit.
      • The Two-Phase Build: Conrad initially built Zenefit rapidly with "no backend" (manual processes) to validate the idea, and only built the complex, feature-rich product after he had earned the right through prior execution.
    • Dropbox and Stripe: While often mentioned as late launchers, these companies also followed a pattern of validating core concepts before scaling, though the specific timeline nuances vary by product complexity.
    • Rarity of the Exception: Founders with the prior context to build robust MVPs immediately are rare; most founders possess only 5% of the necessary story knowledge but attempt to build the remaining 95% before showing the product.
  • Strategic Recommendations

    • The Waitlist Distinction: Creating a waitlist is not a launch; it is merely a form collection that fails to generate actual product feedback or user behavior data.
    • Historical Analysis: Founders should examine early blogs and screenshots of current market leaders (e.g., GitHub, Stripe) to realize their products started with lower stakes, simpler features, and "uglier" interfaces than their current perception of the final product.
    • Conditional Advice: Only founders who have previously built a billion-dollar company in the exact same domain should ignore the "move fast" advice; all others should launch immediately and accept that the first iteration will be imperfect.
    • Decision Framework: Founders are constantly choosing between 10 mph and 100 mph speeds; the optimal path is to choose the faster velocity daily until the product-market fit is proven.