Lessons from Archie Abrams

Lessons from Archie Abrams

Archie Abrams is the VP of Product and Head of Growth at Shopify, and previously scaled product teams at Udemy and Lyft. He rejects standard growth advice by banning strict KPIs for core teams and accepting higher early-stage churn to capture long-term upside. This collection covers how he structures growth organizations and builds durable platforms while trusting user intuition over immediate data.

Part 1: The Hundred-Year Mindset

1. On Time Horizons: "We are building a hundred-year company, which means short-term optimizations are often actively harmful to our ultimate goal." — Source: Lenny's Newsletter
2. On Quick Wins: "If an optimization feels like a quick buck, it probably isn't worth building." — Source: TLDL Transcripts
3. On Sustained Success: "Every decision should be about building a product or service that can withstand the test of time." — Source: Lenny's Newsletter
4. On Future Trade-offs: "Stop chasing immediate lifts that feel good now but might sabotage your merchant's future." — Source: Recall
5. On Legacy: "You don't build a durable platform by hacking the next quarter's numbers; you build it by aligning entirely with the user's long-term survival." — Source: Superme.ai
6. On Institutional Patience: "Keeping experiment holdouts for years allows us to see what actually compounds." — Source: Lenny's Newsletter
7. On False Signals: "Initial lifts can be highly misleading. We've seen many early winners turn into long-term losers." — Source: Lenny's Newsletter
8. On Compound Growth: "Real value creation happens quietly over years, not loudly in a two-week sprint." — Source: Boundaryless
9. On Aligning Incentives: "If you want a team to think in decades, you cannot bonus them based on monthly conversion goals." — Source: Refound AI

Part 2: Rethinking KPIs and Metrics

1. On Banning KPIs: "We actually ban KPIs for the core product team because rigid metrics often limit what people are willing to try." — Source: Lenny's Newsletter
2. On Goal Distraction: "When you give a product manager a specific numerical target, they will hit it, but they might break the user experience to do so." — Source: TLDL Transcripts
3. On Data Dependency: "Short-term data is frequently a trap that convinces you you're making progress when you're just moving numbers around." — Source: Recall
4. On Total Cohort Value: "We focus on total cohort value. We incentivize teams to think about the absolute count of successful merchants, not funnel conversion rates." — Source: Boundaryless
5. On Funnel Constriction: "The easiest way to increase retention is always to constrict the funnel stage one above. But that doesn't mean it's the right thing to do." — Source: TLDL Transcripts
6. On Meaningless Friction: "You can always increase a conversion rate by making the previous step harder. It rarely results in more successful users." — Source: Appetals
7. On Success vs. Metrics: "Our goal is not to improve the sign-up metric. Our goal is to create more thriving independent businesses." — Source: Superme.ai
8. On Avoiding Local Maxima: "Without KPIs dictating every sprint, teams are free to swing for entirely new paradigms rather than just optimizing button colors." — Source: Lenny's Newsletter
9. On True North: "Your north star should be user success, not internal dashboard greenery." — Source: Refound AI
10. On Measurement Limits: "Some of the most valuable things we build cannot be accurately measured in a 14-day window." — Source: Lenny's Newsletter

Part 3: Experimentation and Data Neutrality

1. On Neutral Results: "Don't shy away from shipping experiments that have neutral impacts if your conviction says it is the right path." — Source: Recall
2. On Failed Lifts: "We found that 30% to 40% of experiments showing positive short-term results actually have zero long-term impact." — Source: Lenny's Newsletter
3. On Trusting the Market: "Let the market respond over time. Sometimes a 'loser' experiment yields unexpected value three years down the line." — Source: TLDL Transcripts
4. On Long-Term Holdouts: "Multi-year holdouts are painful to maintain, but they are the only honest way to measure compounding features." — Source: Lenny's Newsletter
5. On Testing Culture: "A robust experimentation culture isn't about proving you are right; it's about uncovering how often you are wrong." — Source: Instapage
6. On Decisiveness: "It's not about making the right decision, it's about making the decision once you have roughly 70% of the necessary information." — Source: Refound AI
7. On Inconclusive Data: "When an experiment is inconclusive, you don't automatically kill it. You can ship it based on pure conviction." — Source: Focused Chaos
8. On Unbiased Learning: "We ship experiments to learn and adapt, not to artificially inflate our weekly updates to leadership." — Source: TLDL Transcripts
9. On Iteration Speed: "Speed matters, but direction matters more. Make sure your A/B tests are actually pointing toward the horizon." — Source: Superme.ai

Part 4: Optimizing for Churn and Success

1. On Intentional Churn: "We intentionally accept high churn among new merchants because we are optimizing for the massive upside of the successful ones." — Source: Liminary
2. On Lowering Barriers: "We want to make it as easy as possible to get started, fully knowing that most new businesses do ultimately fail." — Source: TLDL Transcripts
3. On Early Failure: "High early churn is a feature, not a bug, if it means you allowed a wider top-of-funnel to try their hand at entrepreneurship." — Source: Lenny's Newsletter
4. On Capturing Upside: "A GMV-based monetization model means you can afford massive churn if the winners stay with you for a decade." — Source: Liminary
5. On Emotional Engagement: "We optimize for churn by looking at emotional engagement. If they are engaged emotionally, they will figure out the logistics." — Source: Lenny's Newsletter
6. On Merchant Health: "Focusing on merchant health means accepting that some users should churn because the timing or product isn't right for them yet." — Source: Boundaryless
7. On Business Survival: "The reality of software is that you succeed when your users survive. Everything else is secondary." — Source: Recall
8. On Friction: "Monetary friction, like trial lengths and incentive structures, heavily dictates who stays and who leaves." — Source: Lenny's Newsletter
9. On Pricing Levers: "Discounting is a massive growth lever, but if used incorrectly, it trains your users to wait rather than to act." — Source: Udemy Courses

Part 5: Growth Team Structure and Leadership

1. On Organization: "I split the 600-person growth team into two distinct groups: Growth R&D and Growth Marketing." — Source: Lenny's Newsletter
2. On Customer Support: "We uniquely roll customer support directly into the growth organization because support is a primary retention engine." — Source: Lenny's Newsletter
3. On The CMO Role: In Lenny's episode, Abrams describes Shopify's growth structure as Growth R&D plus Growth Marketing, with customer support inside growth, and discusses why operating without a standalone CMO can keep marketing, product, and growth closer to user outcomes. — Reference: Lenny's Podcast episode on Shopify growth team structure, marketing structure, and the benefits of not having a CMO
4. On Cross-Functional Alignment: "Clarity in team structure ensures that engineering, design, and marketing understand their exact contribution to the overarching strategy." — Source: Recall
5. On Hiring: "Building a growth team requires hiring for intellectual honesty and a willingness to be proven wrong by the market." — Source: Udemy Courses
6. On Getting Things Done: "Our internal 'Get shit done' framework relies on clear ownership rather than consensus." — Source: Lenny's Newsletter
7. On Team Scalability: "As you scale past 300 people, informal communication breaks. You need rigid frameworks for decision rights." — Source: Superme.ai
8. On Growth Mechanics: "Tactical growth hacking expires. Foundational team structure and leadership dictate long-term output." — Source: Growth Talent
9. On Support as Growth: "Treating support as a cost center is a mistake; it is your frontline for identifying churn risks and expansion opportunities." — Source: Lenny's Newsletter

Part 6: Product Intuition Over Rigidity

1. On Taste: "Taste and intuition are valid reasons to ship a feature, even if the immediate dashboard metrics look flat." — Source: Lenny's Newsletter
2. On Founder Vision: "Working with an opinionated founder requires you to balance rigorous testing with respect for pure product vision." — Source: Lenny's Newsletter
3. On Over-Optimization: "You can optimize the joy entirely out of a product if you only listen to the data." — Source: TLDL Transcripts
4. On Conviction: "Conviction is built through deep empathy with the user's struggle, not by staring at a spreadsheet." — Source: Boundaryless
5. On Breaking Rules: Abrams frames Shopify's growth approach as deliberately breaking standard growth playbooks: the team avoids KPI-driven core product goals, accepts early churn, trusts intuition in some neutral tests, and makes decisions through a hundred-year company lens. — Reference: Lenny's Podcast episode on Shopify banning KPIs, optimizing for churn, trusting intuition, and building for a hundred-year horizon
6. On Creativity: "Removing KPIs from the core team is the fastest way to reintroduce creativity into the product development cycle." — Source: Lenny's Newsletter
7. On Qualitative Feedback: "If merchants are telling you something feels wrong, but the numbers look good, the numbers are probably lying to you about the future." — Source: Recall
8. On Core Utility: "No amount of growth tactics will save a product that lacks a fundamentally intuitive core utility." — Source: Superme.ai
9. On Product Debt: "Shipping quickly is fine, but if you abandon intuition, you accumulate product debt that users feel intuitively." — Source: Refound AI

Part 7: Integrating Sales and Product-Led Growth

1. On Sales Alignment: "Integrating sales into a product-led growth model requires making sure the sales team is incentivized by user success, not just closed contracts." — Source: Lenny's Newsletter
2. On Self-Serve: In discussing sales integration at Shopify, Abrams treats sales as part of the product-led growth system rather than a separate rescue motion: sales should fit the self-serve engine and help larger or more complex merchants succeed. — Reference: Lenny's Podcast episode on integrating sales into Shopify's product-led growth model
3. On Hybrid Models: "The most durable businesses eventually figure out how to layer enterprise sales cleanly on top of a grassroots PLG motion." — Source: TLDL Transcripts
4. On Funnel Ownership: "Sales and product cannot operate in silos; they must share the exact same definition of a qualified, successful user." — Source: Boundaryless
5. On Lead Quality: "If product feeds garbage leads to sales just to hit a conversion metric, the entire system breaks down." — Source: Recall
6. On Sales Feedback: "Sales teams hear the objections that users are too polite to type into a feedback form. Product must listen to them." — Source: Instapage
7. On Enterprise Expansion: "Moving upmarket requires changing the product's administrative capabilities without ruining the simplicity that got you there." — Source: Superme.ai
8. On Growth Synergy: "When marketing, product, and sales report through a unified growth structure, you eliminate the classic lead-attribution turf wars." — Source: Lenny's Newsletter
9. On Monetization: "You have to be careful that your sales motion doesn't cannibalize the trust you built through your self-serve product." — Source: Refound AI

Part 8: The Fundamentals of Inception and Distribution

1. On Defining Growth: "Growth can be explained using four words: problem, inception, distribution, and word-of-mouth." — Source: Medium
2. On Inception: "Inception is the act of planting the idea in the user's mind that your specific product solves their specific problem." — Source: Medium
3. On The Core Problem: "If you cannot clearly articulate the 'why should anyone use it' problem, no distribution hack will save you." — Source: Medium
4. On Word-of-Mouth: "Word-of-mouth remains the only truly scalable acquisition channel that lowers your blended customer acquisition cost over time." — Source: Instapage
5. On Distribution: "Building it is only half the battle. If you don't have a distinct distribution advantage, you are renting your users." — Source: Superme.ai
6. On Paid Acquisition: "Scaling paid channels works until it doesn't. It must always be subsidized by a strong organic baseline." — Source: Udemy Courses
7. On SEO and Content: "Organic content and SEO are long-term investments that act as a moat against rising paid media costs." — Source: Instapage
8. On Marketplace Dynamics: Abrams explains in the Instapage transcript that Udemy handled growth differently on each side of the marketplace: the student side was a growth marketing and growth product problem, while the instructor side was more about content quality and instructor community management. — Reference: Instapage transcript with Abrams on Udemy separating student demand growth from instructor-side quality and community work
9. On Consumer Habit: "At Lyft, the focus was entirely on turning an occasional utility into a deeply ingrained consumer habit." — Source: Superme.ai
10. On Education: "Educational marketing is powerful because it gives the user value before you ever ask them for a credit card." — Source: Udemy Courses
11. On Retention Realities: "Acquisition without retention is just a very expensive way to rent traffic." — Source: Recall