Mike Duboe: Top 5 Lessons Scaling Stitch Fix to IPO; Why CAC/LTV is a BS Metric | 20VC #978

Growth Philosophy & Mindset

• Holistic Objective Functions: Growth teams must optimize for down-funnel metrics to prevent siloed behavior where acquisition teams degrade retention (e.g., bringing in low-quality users).
• Incrementality over Attribution: Relying on last-click or fractional attribution models creates "false precision"; the most accurate measurement of marketing efficacy is holdout testing to measure lift.
• Growth vs. Growth Marketing: Paid marketing is an "accelerant, not a crutch"; heavy reliance on paid channels too early masks product-market fit issues as costs degrade with scale.
• Loops vs. Funnels: Operationalizing growth via "loops" (where inputs generate outputs that reinvest into inputs, e.g., Pinterest's content creation) creates compounding growth, whereas linear "funnels" (AARRR) often lead to siloed teams and decaying performance.
• Definition of Head of Growth: The role is responsible for (1) accelerating the company's pace of learning via experiments and (2) engineering systems to build control over North Star metrics.

Organizational Structure & Timing

• Reporting Lines: Sustainable growth programs requiring engineering discipline should sit under Product; marketing-led growth is often suitable only for simple, linear e-commerce funnels.
• CEO Reporting: At scale, a growth team often reports to the CEO to maintain the autonomy required to run aggressive tactics and bypass engineering/product dependencies.
• Cross-Functional Pods: Preferred structure involves embedding product managers, engineers, designers, and analysts into autonomous pods to run experiments without blocking the core roadmap.
• Hiring Timing: Founders should never hire a growth lead pre-product-market fit; early hires should be generalists focused on user discovery rather than scaling top-of-funnel.
• First Hire Role: The first growth hire should be an analytics professional (not a data scientist) to eliminate "analytics debt" and establish data instrumentation foundations.

Hiring & Interview Framework

• Defining Success: Before posting a job, founders must define specific outcomes (e.g., "build a growth model by month X") rather than vague mandates to "grow the business."
• Reference Checks: Conduct reference checks early in the process to validate risk appetite and cultural fit, rather than treating them as a final formality.
• Key Interview Questions:
  • "Pick your favorite product and explain how it grows" (tests holistic system thinking).
  • "Tell me about a time you failed" (tests willingness to iterate and learn from experimentation).
  • "Show me your daily dashboard" (tests ability to prioritize high-signal metrics over noise).
• Take-Home Exercises:
  • Experiment Roadmap: Request a prioritized list of experiments with ROI context to assess triage skills (avoiding the "test everything" mindset).
  • Growth Model: Ask candidates to write a growth model for a specific product relevant to the business to test systemic visualization skills.
• Compensation Bands (Series B):
  • Base Salary: $150,000 – $300,000.
  • Equity: 0.5% – 1.0% (higher equity recommended to retain "barrel" talent who drive compounding output).

Experimentation & Analytics Infrastructure

• Experiment One-Pager: A standardized framework for all experiments including objective, hypothesis, design/resources, timeline, quantifiable success criteria, and next steps.
• Triage System: Use a centralized spreadsheet (e.g., Google Sheet) for company-wide visibility, utilizing the ICE framework (Impact, Confidence, Effort) to prioritize ideas.
• Learning Codification:
  • Weekly Reviews: Growth team meetings to review completed experiments and lessons learned.
  • All-Hands Sharing: Communicating key learnings (including invalidated hypotheses) to the wider org to prevent "reinventing the wheel" and align product strategy.
• Analytics Debt Indicators: High debt exists if answering basic cohort questions (e.g., Day 30 retention) requires custom database queries taking days, or if product instrumentation lacks key user event logging.

Paid Marketing Strategy

• Budget Allocation: Paid spend should be driven by payback period thresholds (e.g., first-transaction payback for e-commerce, 12-month payback for subscriptions) rather than arbitrary percentages.
• Channel Diversification: Avoid >50% concentration in any single channel at scale to mitigate risk (e.g., policy changes or platform performance drops), though early-stage teams should double down on a single winning channel.
• Creative Volume: Sustainable advantage in performance marketing comes from high-volume, low-production creative testing to feed algorithms, rather than polished assets.
• Incrementality Testing: Even for "brand" channels like TV, use holdout tests (e.g., pausing national ads while running local spots in test regions) to calculate true incremental CPA and lift.
• Critique of LTV:CAC: The LTV:CAC ratio is flawed due to the unknown "lifetime" in early stages and lack of granularity; payback period is a superior metric as it avoids lifetime assumptions and allows for channel/keyword-level analysis.

Forward-Looking Statements & Trends

• Future of Growth: The discipline of growth will continue to accelerate as product and marketing leaders internalize these principles, moving away from the title-centric "growth team" model.
• Fair's Referral Mechanic: Highlighted as the most impressive recent strategy; a B2B referral engine that drove sustainable, compounding growth through cross-side network effects.
• Over-Incentivization Risk: Early experience at Tilt showed that incentivizing usage too heavily creates dependency, making it difficult to transition users to non-incentivized organic behavior.