Breaking the rules of growth: Shopify's unconventional playbook

Executive overview

Shopify optimizes for enabling entrepreneurship over retention, accepting that most businesses will fail but allowing the rare winners to generate exceptional returns. The company bans KPIs for core product teams, making decisions based on founder taste and intuition aligned with a 100-year vision. Because the growth team measures long-term GMV impact—often waiting 1–3 years to see true effects—they've discovered that 30–40% of short-term wins actually produce no incremental lift downstream.

Core insight: Optimizing local metrics (conversion rates, retention) backfires because teams naturally restrict the upstream funnel. Measuring absolute numbers of successful merchants and long-term cohort value forces teams to expand opportunity rather than narrow it.

Why Shopify optimizes for churn, not retention

• Most new merchants fail, but winners produce extraordinary GMV through the payment model
• Revenue is tied to merchant success (payments, not subscription alone), so high-value winners offset many losses
• Lowering barriers to entry increases total cohort value even if short-term churn is higher
• Think of it like angel investing: most bets fail, but outliers drive portfolio returns

The three-week test, one-year hold

• Every experiment ships after three weeks, but the exposed cohort is tracked for 12+ months
• Automated alerts show teams the actual long-term impact at three, six, nine, and 12 months
• Prevents premature celebration and forces accountability to real outcomes
• 30–40% of experiments show no long-term lift; another subset flips negative or positive downstream

Why short-term wins often vanish

• Payment failure notifications looked like major lift, but selection bias meant those merchants weren't committed anyway
• Lower-quality users pulled forward from future cohorts, not genuinely acquired
• Many discount experiments didn't create new entrepreneurs; they just shifted purchase timing
• Exception: Pre-configured store sections showed no signup lift but massive GMV impact later as merchants built better stores, got early sales, and stuck with the platform

Absolute numbers over conversion rates

• Teams naturally optimize local conversion rates (my step in the funnel) because it's easiest
• But increasing signup-to-activation rate is trivial if you just make signup harder
• Shift incentives to total merchants reaching each stage, not rates
• Making signup easier hurts conversion rate but often drives more absolute activated merchants

Monetary friction as a core lever

• Trial length, trial amount, discounts, and app store credits all reduce barriers for unproven merchants
• Counterintuitive: small discounts unlock a class of merchants who would've given up, giving them time to prove the business
• Higher-intent merchants aren't better than lower-friction merchants; they're just fewer
• Bootstrapping founders need runway; reduce friction and they often become high-value sellers

Three-part growth organization

Growth R&D: Product, design, engineering, data, tooling

• Growth products: signup, onboarding, monetization, engagement across the full merchant lifecycle
• Enable pillar: experimentation platform, communication tools, BI, marketing tech
• Customer support team: tooling and self-serve AI to help new merchants succeed

Growth marketing: Paid channels, SEO, email, content, affiliate partnerships; each channel operates under LTV guardrails

Measured by cohort value, not rates

• Macro goal: total GMV from a cohort over three to four years
• Include payback guardrails for efficiency
• Each sub-team measured on incremental cohort value lift from experiments
• Irrelevant whether conversion rate went up or down; relevant whether merchants produced more GMV long-term

Why core product doesn't use KPIs

• Banned OKRs and metrics; decisions driven by taste, intuition, and long-term architecture
• Technical "how" determines future strategy more than "what" or "who"
• Every project requires OK2 approval from Glenn (head of core product) and others before shipping
• Twice-yearly big releases; continuous smaller shipping, but all gated by quality bar

The tension between taste and metrics

• Growth team optimizes fast, core product prioritizes technical foundation
• Resolved by trust, commitment to quality, and willingness to run experiments
• Example: No-Wizard principle prevents easy onboarding carousels; instead, pre-filled sections baked into the editor itself
• Healthy conflict requires respecting each side's constraints

The hundred-year vision

• Toby (CEO) drives all decisions back to: What's right for commerce 100 years from now?
• Short-term revenue (acquiring existing big brands) is seductive but doesn't build the platform
• Better to enable every new entrepreneur, even if most fail, because future winners were never built-in to today's forecast
• Technical decisions reflect optionality; avoid lock-in even if it means slower shipping

Sales and self-service hybrid funnel

• Historically product-driven; now layering in sales motion for enterprise
• Growth is building tools (CSV importers, guidance) for sales-driven onboarding
• Broke traditional attribution by merging self-serve and sales LTV; had to rebuild all measurements
• Moving toward incrementality testing (causality) rather than multi-touch attribution

Onboarding and guidance consistently outperform

• Collecting the right information in signup, personalizing guidance
• Surprising that simple execution of best practices drives outsized impact
• Investing in the first experience sets merchants up for long-term success
• Often the biggest lever, even for a company that doesn't prioritize retention

Discounting lessons from Udemy

• High list price signals quality; discount signals affordability
• Urgency + discount taps into emotional job of "I'm making educational progress"
• Purchasing is progress, even before consumption
• Education is aspirational; the act of buying changes behavior downstream

No CMO; marketing embedded

• Growth marketing in growth org, revenue marketing near sales, brand under Harley, PMM with product teams
• Benefit: closest to primary outcomes; move faster with less coordination
• Requires founder (Toby) and COO (Harley) with strong brand intuition
• Downside: messy, subjective, needs senior people to hold the bar