Know customers so well you can predict their responses
The standard for customer understanding: speak with customers so much that you start predicting what they're going to say next, because you've heard it so many times before.
When to use
When evaluating customer understanding; when seeking PMF
Don't do this
Assuming you understand customers without deep conversation
17 Founders Who Did This
Focus on customer value, not competitors - track the industry but don't obsess
Stay in the 'feedback trenches' longer than feels comfortable
Listen to users continuously and ship fixes within 24 hours
Invest time learning about customers rather than just product features
Relentlessly stay in front of your customers to truly understand their needs and pain points
Ignore feedback from users who don't love your core value—focus on "somewhat disappointed" users for whom your main benefit resonates
Customer support is the most important startup function at any stage
Used pain-point messaging 'good at lawns, not paperwork' instead of feature lists in cold emails to lawn care businesses
Larson recommends 'conflict mining' - finding controversial topics and having conversations with skeptical ICs to uncover context. When an IC resisted his proposal at Stripe, deeper discussion revealed he was missing critical context about core architecture constraints.
Built features based solely on customer emails and dropoff analytics - fixed complaints, improved problem pages, removed unused features
Called 3,000 customers before building cooling blanket. Customers all said they would buy it if made. Aaron reached the point where he could predict customer responses because he had heard them so many times.
Developed heuristic after failed products: keep having customer conversations until three-quarters of the content is stuff you already knew, indicating deep enough market understanding to build
Practiced curiosity-driven pattern matching to understand individual engineer motivations, categorizing team members as compensation-driven, challenge-seeking, or progression-focused through repeated observation
Adopted extreme ownership philosophy where any blocker to customer success was Mutiny's responsibility. Built integrated analytics, visual editors, and automated recommendations to remove friction points rather than expecting customers to find workarounds.
Uses multiple independent perspectives (personal analysis, team members, contractors, AI) to review customer interview recordings, reducing cognitive bias in interpreting feedback
Spoke with customers so extensively that he aimed to predict what they would say before they opened their mouth
Does the role himself before hiring to understand seniority needed, biases toward 'no' when on fence, spends 40 minutes of interviews on candidate motivations not skills