Eric Migicovsky - How to Talk to Users

Core Philosophy and Strategic Imperatives

• Founders at Y Combinator (YC) must maintain a direct, personal connection to users throughout the entire company lifecycle to extract critical data; this cannot be outsourced to sales or product teams.
• The two fundamental requirements for starting a company at YC are coding/building the product and talking to users.
• The "go-between" role between engineers and users is explicitly discouraged; founders and core technical members must develop user interview skills to avoid reliance on intermediaries.
• Successful user interviews focus on extracting data about the user's life and problems, not pitching the product or selling the founder's vision.
• The presentation synthesizes advice from the book The Mom Test, identifying three common errors in user interviews: pitching the idea, asking hypothetical questions, and talking too much.

Tactical Interview Methodologies

• Avoid Pitching: Do not present the product idea during an interview; the goal is to learn about the user's existing life and specific problem areas.
• Avoid Hypotheticals: Do not ask "If we built feature X, would you use it?" because users are poor predictors of future behavior; instead, focus on specific past events.
• Listen More: Founders should minimize talking and prioritize listening to capture hard data about user lives to bring back to co-founders.
• Five Critical Interview Questions:
  • "What is the hardest part about doing the thing you are trying to solve?" (Validates if the problem is a genuine pain point).
  • "Tell me about the last time you encountered this problem." (Extracts specific context, timing, and circumstances of the problem).
  • "Why was this hard?" (Reveals the underlying motivations and specific friction points that inform marketing and sales copy).
  • "What have you done to try to solve this problem?" (Determines if the problem is "burning" enough to warrant a solution and identifies existing competitors).
  • "What don't you love about the solutions you've already tried?" (Identifies feature gaps and differentiators without asking users to imagine future features).

Customer Acquisition and Identification Strategies

• Finding First Users:
  • Start with oneself as the user to test the interview strategy.
  • Leverage warm introductions through friends and co-workers.
  • Use "guerrilla" tactics like showing up in person (e.g., visiting fire stations) rather than relying on cold email.
  • Attend industry events without booths to meet potential users in informal settings like coffee shops.
• Prototype Stage Framework:
  • Identify the "best first customer" by quantifying three specific variables:
    • Cost: How much money does the problem cost the user daily (revenue loss or expense)?
    • Frequency: How often does the problem occur (daily vs. yearly)? Frequent problems create more engagement.
    • Budget/Authority: Does the user have the budget and decision-making power to solve the problem?
  • Visualize these three variables as overlapping Venn diagrams to pinpoint the ideal customer profile.
  • Case study: A hypothetical smart blender found McDonald's to be the best first customer due to high volume and budget authority, despite lower per-transaction value compared to high-end restaurants like The French Laundry.

Iterating Toward Product-Market Fit (PMF)

• Measuring PMF Quantitatively:
  • Vague definitions like "people want it" are lagging indicators; quantitative metrics are needed for real-time iteration.
  • Reference Superhuman's "40% Rule": Ask users weekly, "How would you feel if you could no longer use this product?"
  • If 40% or more users answer "Very Disappointed," the product has reached product-market fit, predicting exponential growth.
  • Companies with PMF consistently report this 40% threshold when evaluated against other successful firms.
• Tactical Execution for PMF:
  • Capture Contact Info: Request phone numbers during sign-up to enable direct follow-up on specific user data trends.
  • Validate via Action, Not Questions: Instead of asking if a user wants a feature, ask them to enter credit card information or pay for an upgrade before the feature is built to test genuine demand.
  • Discard Bad Data: Ignore compliments ("I love the design") and hypotheticals ("In the future I would want...") as they do not provide actionable insights for feature prioritization.
  • Focus strictly on specific past behaviors and concrete problems to avoid designing "by committee."

Historical Context and Founder Credibility

• Eric Mitryakovsky is a YC partner and the founder of Pebble (2011), the creator of one of the first smartwatches.
• Pebble's strategy included attending CES without a booth to meet users in coffee shops, demonstrating low-budget customer acquisition tactics.
• The presentation emphasizes that founders should take notes immediately or record interviews, as raw data quality is often forgotten later.
• Initial interviews should be casual and short (10–15 minutes), requiring no rigid scheduling or large sample sizes to begin.