Founder Habits You Need To Drop To Be A Great CEO

Key Concepts

• AI Reinvention of Analytics: The core idea that Artificial Intelligence will fundamentally change how analytics platforms operate and are used.
• Amplitude's AI Strategy: Amplitude's journey from AI skepticism to actively integrating AI into its product offerings, aiming to lead this transformation.
• Challenges of AI Adoption for Incumbents: Difficulties faced by established companies in reorienting and rebuilding to effectively leverage AI, often due to existing engineering mindsets and organizational inertia.
• "Grumpy Engineers" and AI Skepticism: The phenomenon where engineers, particularly in established tech companies, may be skeptical of AI's capabilities or resist its adoption.
• "AI Native" vs. "SAS Native" Mindsets: The difference in approach and understanding between engineers who grew up with AI tools and those accustomed to traditional Software-as-a-Service (SaaS) development.
• Technology-First vs. Customer-First Approach in AI: The necessity of understanding AI capabilities first before mapping them to customer needs, contrasting with the traditional SaaS model of asking customers what they want.
• Top-Down vs. Bottom-Up AI Adoption: The observation that AI adoption in many companies is driven by leadership (top-down) rather than engineers (bottom-up), contrary to typical technology adoption patterns.
• Founder Mentality vs. Executive Role: The shift in responsibilities and skillsets required when transitioning from a startup founder to a large company executive.
• Intrinsic Motivation and Startup Resilience: The critical role of a clear "why" and internal drive in overcoming the inevitable challenges and doubts faced during the early stages of a startup.
• Commoditization of AI Visibility: The idea that basic AI-powered visibility tools will become easily accessible and free, necessitating a focus on downstream business applications.
• The "Cursor Moment" in Analytics: The prediction that AI will create a paradigm shift in how analytics are used, similar to how Cursor changed code editing.
• The Importance of a Clear Mission and Goal Tree: The strategy of defining a core mission and then breaking it down into actionable goals to maintain focus and direction.

Amplitude's AI Transformation and the Future of Analytics

This discussion with Spencer Skates, CEO and co-founder of Amplitude, delves into the company's strategic pivot towards Artificial Intelligence and its implications for the analytics industry. Amplitude, a leading analytics platform used by major companies like Curser, Door Dash, and Walmart, has navigated a significant transformation from initial AI skepticism to becoming a leader in AI-driven analytics.

1. Amplitude's Journey with AI: From Skepticism to Leadership

• Initial Hesitation: Amplitude, like many established companies, was initially skeptical of AI. The relevance of AI began to grow in 2022-2023, but Amplitude didn't actively pursue it until late 2024, recognizing its potential to reshape the analytics landscape.
• Reasons for Skepticism:
  • "Grifting" in AI: Co-founder Jeffrey was particularly frustrated by what he perceived as hype and unrealistic promises surrounding AI, such as job replacement and a world of abundance, without clear practical applications.
  • Jagged Capabilities: The early AI models had very uneven capabilities, excelling in some areas while being "absolutely terrible" in others. This made it difficult for those without deep technical understanding to provide actionable guidance.
  • Misguided "AI Strategy" Discussions: Executives and investors often asked about an "AI strategy" without a clear understanding of what that entailed, leading to frustration.
• The Turning Point: The transformative effect of AI on software engineering, particularly with tools like Cursor, Claude, and others, demonstrated increased productivity. This was a key indicator for Amplitude that "there's something there."
• Strategic Shift (October 2024):
  • Hiring Wade Chambers: A new engineering leader with extensive AI experience was brought in.
  • Acquisition of Command AI: A YC company specializing in AI-driven user guidance and chatbots was acquired. Command AI was developing a product that intelligently triggered guides for confused end-users and featured a chatbot similar to Intercom's Finn.
• Product Launches and Future Vision: Amplitude has recently launched AI Feedback, AI Visibility, and its MCP server. They plan a significant expansion in December-January-February with "Cursor for Analytics," aiming to dramatically change how people use and leverage analytics.

2. Internal Dynamics and AI Adoption Challenges

• Organizational Inertia: Reorienting and rebuilding a larger company to use AI effectively is challenging. Amplitude, with around 800 employees (200 in product, engineering, and design), found it took a full year to get the team fully on board and believing in AI.
• Investor and Board Pressure: Following the public launch of ChatGPT and the surge in AI discussions, investors and board members began inquiring about Amplitude's AI strategy, often driven by external media coverage.
• Internal Team Sentiment: While a few individuals were experimenting with AI, the broader team was focused on existing product roadmaps, including experimentation, session replay, and activation features. The massive potential shift of AI was not yet fully recognized organizationally.
• The Role of Change Agents: Wade Chambers and the Command AI team were crucial "change agents" in demonstrating AI's possibilities and gaining organizational buy-in.
• "AI Week" as a Pivot Point: A key initiative was an "AI Week" (delayed to June) designed to train leaders and the broader team on AI capabilities. This involved:
  • Leader Adoption: VPs of product and engineering managers used AI tools to understand their potential.
  • Demonstration: A product leader demonstrated a "dark mode" for Amplitude, showcasing AI's capabilities and even troubleshooting a bug live. This visual proof was impactful for the entire engineering, product, and design organization.
  • Hackathon: The rest of the week involved a hackathon where teams used AI tools like Cursor to accelerate existing work.
• Step One: Tool Adoption and Buy-in: The initial focus was not on building specific AI products but on getting the existing team to use and believe in AI tools.

3. The Fundamental Difference: SAS vs. AI-Native Product Development

• Traditional SaaS Model:
  • Customer-Centric: Go to customers, ask what they want and will pay for, prioritize, build, and iterate. This loop is Amplitude's mastered competitive advantage.
  • Clear Requirements: Customers can articulate their needs for existing workflows.
• AI-Native Product Development:
  • Technology-First: Requires familiarity with AI model capabilities to understand what is possible.
  • Customer Inability to Articulate Needs: Customers cannot easily describe what AI can do for them, leading to requests like "give me a faster horse" or vague desires for insight delivery.
  • Mapping Capabilities to Product: The key is to map AI capabilities back to the product's function.

4. The Top-Down Influence on AI Adoption

• Sam Altman as a Salesperson: Spencer attributes the top-down AI adoption trend partly to Sam Altman's effective communication of an ambitious vision for OpenAI, convincing executives, investors, and world leaders of AI's impact.
• Aspiration vs. Capability Gap: This has created a disconnect where there's significant desire for AI solutions, but the underlying capabilities are still catching up.
• Engineer Frustration: This gap leads to frustration among engineers who perceive "grifting" and a lack of tangible execution in the AI space.
• Amplitude's Case Study: Amplitude experienced this frustration internally until the capabilities became more apparent and actionable in the past year.

5. Founder-Led Transformation and Organizational Change

• Founder Mode: Spencer emphasizes the need for a founder-led approach to drive transformation, drawing parallels to how founders lead significant changes.
• Learning by Doing: The best way to understand AI's possibilities is to use the technology and be on the front lines. This necessitates a bottoms-up approach to training the organization.
• Emergent AI Products: Many of Amplitude's AI initiatives, like the MCP server and AI Visibility, emerged organically from engineers' excitement and exploration.
• AI Visibility's Impact: The free AI Visibility product, initially developed by engineer Leo Jen, led to a doubling of new sign-ups to Amplitude's free plan.
• "Ask AI" (Future Launch): A global chat interface, similar to Cursor, will allow users to interact with AI to pull charts, perform analysis, and understand data anomalies.
• Organizational Reorganization: Amplitude has undergone two reorganizations in the engineering, product, and design departments this year to align with its AI-centric future. This involved moving out individuals who were not a good fit for the new direction.
• Acquisitions: Amplitude has acquired several companies, including Craftful, Anari, and June, integrating their founders and teams.

6. The "AI Native" Engineer vs. The "SAS Native" Engineer

• Mentality Shift: The difference is not strictly age-based but a fundamental shift in mentality.
• SAS Native Engineers: Accustomed to the customer-request-build loop. They may struggle to envision entirely new possibilities driven by AI.
• AI Native Engineers: Naturally adept at leveraging AI tools and understanding their potential.
• The "Why" in AI: AI-native teams sometimes lack the deep understanding of existing product and problem-solving expertise, leading to reinventing the wheel.
• Amplitude's Adaptation: Successful engineers at Amplitude have recognized that code is a means to an end (solving customer problems) and have embraced new technologies like AI to achieve this.

7. The Nuance of AI and SaaS: Beyond "AI Killing SaaS"

• High Guarantees in Business Workflows: For many business workflows, especially in areas like CRM, high guarantees of performance and data integrity are fundamental. SaaS excels at providing these.
• Overshooting the Mark with Agents: Some AI companies are trying to have agents handle entire workflows end-to-end, which may be premature.
• Importance of Editing and Redoing: The ability to edit and refine AI outputs is crucial, as highlighted by Gary's experience with AI agents. Creating products that facilitate this is a key challenge.

8. Resource Allocation: Balancing AI Bets and Existing Products

• Integrated Approach: AI products are not entirely separate; the goal of "Ask AI" is to make the existing Amplitude product easier to use.
• Four Key Priorities for Next Year:
  1. Rebuild Amplitude to be AI-native.
  2. Make the product significantly easier to use.
  3. Ensure other products (outside analytics) have competitive parity.
  4. Serve marketers effectively to compete with legacy martech companies.
• Continued Development of Core Products: Foundational work continues on session replay, experimentation, and guides/surveys. For example, "zoning" for session replay will overlay analytics on web pages.
• Increased Productivity: The team's embrace of AI has made them faster and more productive, leading to more output.
• Dedicated AI Teams: While a dedicated team was formed for AI projects, existing teams continue to work on improving the core Amplitude product.
• Self-Selection and Culture: Employees naturally self-selected into AI-focused work. The company communicated a "burning the boats" mentality, signifying a full commitment to AI. Designers, like Will Newton, also gravitated towards AI initiatives.

9. The "AI Visibility" Debate and Business Models

• "Features, Not Companies" Argument: The idea that AI visibility tools are mere features, not standalone companies, was a point of contention.
• Commoditization of Visibility: Spencer argues that AI visibility is easily replicable and will become commoditized and free, as Amplitude has demonstrated by offering its AI Visibility product at no cost.
• Downstream Business Models: Real businesses will be built downstream from AI visibility, focusing on content generation (like Aerops) or other value-added services.
• Amplitude's Advantage: With a substantial existing revenue base, Amplitude can afford to offer AI visibility for free, driving lead generation.

10. Identifying Vulnerable Markets and Competitive Advantages

• Google as a Target: Spencer identifies Google as a prime example of an incumbent that is slow and conservative in B2B, creating opportunities for disruption in areas like email and workspace tools (e.g., Notion competing with Google Docs).
• "Cursor Moment" in Analytics: He predicts a similar paradigm shift for analytics with AI, making current methods seem obsolete.
• Niche Focus: The success of AI businesses will likely come from focusing on specific problems and buyers, rather than generalized agent builders.
• Enterprise AI Adoption Challenges: Security and compliance concerns are major hurdles for enterprise AI adoption, presenting an opportunity for companies that can address these issues.
• "Uber for Tech Support": A persistent market gap exists for a scaled solution to connect tech-savvy youth with older individuals needing IT support.

11. The Genesis of Amplitude: From Voice Recognition to Analytics

• Sonet Light (Voice Recognition): Amplitude's founders' first venture was a voice recognition company, an early precursor to Siri. Despite a strong demo and press, the product and technology were not good enough.
• In-House Analytics: During their time at Sonet Light, they built their own analytics tools because they believed it was the right way to build products by understanding user behavior.
• Pivot to Amplitude (June 2012): After shutting down Sonet Light, they showed their in-house analytics to other companies, who expressed interest. This led to a pivot to Amplitude, focusing on analytics.
• Crowded Analytics Market: Analytics was a highly competitive space, but Amplitude differentiated itself by focusing on a problem they felt uniquely suited to solve with their algorithms engineering background.
• Learning to Sell: Spencer recognized a gap in founders' ability to get technology into customers' hands and sell it, making this a personal learning objective.

12. Learning and Mentorship: The Art of Sales and Finding Guidance

• Imposter Syndrome in Sales: Spencer admits to still feeling like an imposter in sales despite his success.
• Sales Motion vs. Single-Player Mode: Some products (like Cursor, Slack) can be adopted individually, while others require convincing multiple stakeholders, necessitating a sales motion.
• Learning Through Doing and Coaching: Sales is not learned from books but through practice and expert coaching.
• Mitch Mirando as a Mentor: Amplitude worked with Mitch Mirando, a sales executive and coach, who provided rigorous guidance, pushing Spencer to understand customer pain beyond just wanting dashboards.
• Finding Mentors: The best advice is to be clear about what you want to learn and be open to where that guidance comes from.

13. The Founder's Journey: Mission, Resilience, and Leadership Evolution

• Defining a Career Mission: Before starting a company, Spencer clarified his career goal: to dedicate himself to a mission greater than himself and contribute to humanity through building and selling software.
• Goal Tree: This overarching mission is broken down into a "goal tree" (product building, selling, etc.).
• The Emotional Toll of Startups: Starting a company is emotionally painful, with frequent moments where quitting seems rational.
• The "Why" as an Anchor: Returning to the original "why" is crucial for overcoming these difficult periods.
• The "Rational Point to Quit": A key takeaway from "Founders at Work" is that successful founders often persist past the point where quitting seems rational, typically around one to two years in. This resilience is a primary filtering criterion.
• Intrinsic Motivation: Lasting through challenges requires intrinsic motivation, not external validation or financial gain.
• Evolution to Large Company Executive:
  • Founder's Role: Running towards the most difficult problems and leading from the front.
  • Executive's Role: Cannot lead by example everywhere due to the breadth of responsibilities (sales, marketing, product, customers, press). Requires disciplined time management and saying "no."
  • Unlearning Habits: Executives must embrace the role of judging others' work, a shift from the founder's hands-on approach.
  • Hierarchy: Hierarchy is necessary for ownership and team leadership.
  • Leverage and Resources: Large company executives have significant resources and leverage, working less hard but needing to deploy resources effectively. This transition is the hardest.
• Nuance in Management: Management is learned through experience and coaching, not solely from books.

14. Amplitude's Future and Sharing Knowledge

• Playbook for SAS Reinvention: Amplitude's journey is seen as a potential playbook for how SaaS companies can reinvent themselves in the AI era.
• Judicious Time Management: As a public company CEO, time is a critical resource that must be managed carefully.
• Vocal Leadership: Spencer is deliberately more vocal and public with his thoughts on Twitter and other platforms to share his story and insights, encouraging others to learn.
• Authenticity: He aims to express his genuine self and convictions, rather than adopting a persona or engaging in unproductive arguments.
• Leading the AI Reinvention of Analytics: Amplitude's ultimate goal is to lead the upcoming reinvention of analytics through AI.