The Lean Startup | Eric Ries | Talks at Google

Eric Ries introduces himself as the founder of the Lean Startup movement, currently writing a book on the subject. He comes from a programming background and has observed that new product development often wastes time and energy. The Lean Startup offers a solution through five core principles, which he will discuss, and encourages audience interaction.

Ries emphasizes that entrepreneurship is a form of management, not limited to 'two guys in a garage,' but rather a distinct kind of work measured by validated learning. He contrasts this with traditional management and the romanticized stories of entrepreneurship, like 'The Social Network,' which focus on the idea and the outcome, largely ignoring the core work of figuring out what customers want and prioritizing features—the 'photo montage' phase.

A startup is defined as a 'human institution designed to create something new under conditions of extreme uncertainty,' regardless of size or industry. Ries argues that startups are experiments, akin to scientific endeavors with hypotheses and predictions. The biggest waste in product development is efficiently building something nobody wants, as evidenced by the high failure rate of Web 2.0 companies, which often fail despite being technically feasible.

Ries links traditional management practices to Frederick Winslow Taylor's scientific management, which, while revolutionary for its time, treats workers as automatons and assumes predictable outcomes. He argues that this 'great man theory' of management is still applied to knowledge work and entrepreneurship, leading to inefficient new product development. He suggests the need for a new management paradigm specifically for entrepreneurship.

Ries introduces the concept of a 'pivot'—a fundamental change in strategy without a change in vision—as a universal constant in successful startups. He argues that successful founders don't necessarily have better initial ideas but are adept at pivoting. The goal is to reduce the time between pivots to increase the odds of success before running out of resources. This requires focusing on 'validated learning' rather than blindly following a plan.

Ries criticizes the 'waterfall' methodology for software development, calling it an obsolete model. He introduces the 'build-measure-learn' feedback loop, inspired by lean manufacturing, as a cyclical process for turning ideas into code, measuring customer interaction, and learning. He illustrates this with his experience at IMVU, where their initial strategy for a 3D avatar instant messaging add-on was flawed because customers didn't want to use it to connect with existing friends, leading to a necessary pivot.

The IMVU experience highlighted that much of their initial code was 'waste' because it didn't contribute to validated learning. Ries emphasizes that startups exist to learn how to build a sustainable business, and a 'minimum viable product' (MVP) should contain only what is necessary to learn if a plan is correct. The heuristic for evaluating any startup advice is whether it minimizes the total time through the build-measure-learn loop.

Ries introduces 'innovation accounting' as a new system for holding entrepreneurs accountable, moving away from 'vanity metrics' that don't reflect actual progress. Instead, he advocates for 'actionable metrics' that focus on per-customer behaviors to establish a baseline. Product development then focuses on 'tuning the engine' to improve these metrics, rather than just building features. The ultimate decision to pivot or persevere should be based on this quantitative data.

In the Q&A, Ries addresses how much effort to put into a first pivot, stressing the importance of making specific, concrete predictions to determine if a strategy is working. He also discusses the applicability of Lean Startup principles to enterprise applications, arguing that while tactics may differ (e.g., smaller customer bases allow for higher-fidelity experiments), the core build-measure-learn principles remain universal.

Ries advises Google against putting its brand name on initial, risky products. He recounts an instance where Google launched a product that lagged behind IMVU's two-year-old version, and the public launch pressure prevented Google from pivoting effectively. He argues that obscurity is a benefit for early-stage products, allowing teams to experiment and fail quickly without public embarrassment. Ries suggests that leadership should create safe environments for experimentation and celebrate successful pivots, rather than just failures or 'success theater.'