The business world has long grappled with the optimal approach to product development, traditionally navigating between at least two distinct methodologies. For decades, the predominant method, even persisting into the current era of artificial intelligence, has been the project model. This approach is fundamentally concerned with output, wherein executives and stakeholders delineate a prioritized roadmap of features and projects. Subsequently, a product manager, often functioning as a "feature team product manager," drafts a detailed specification, such as a Product Requirements Document (PRD). This specification then informs a designer’s creative work, and ultimately, engineers are tasked with building precisely to that design and specification. This sequential, directive-driven process, while seemingly structured, has increasingly revealed its limitations, often leading to what industry experts term a "feature factory"—a system capable of rapidly producing numerous products that fail to meet market needs or deliver genuine value.
In contrast, an alternative paradigm, known as the product model, places its emphasis firmly on outcomes. This model empowers product leaders to identify significant problems requiring solutions. These problems are then tackled by autonomous, cross-functional product teams. Their initial mandate is not to build, but to discover a viable solution, gathering conclusive evidence that the proposed solution can indeed achieve the desired outcome. Only after this rigorous discovery phase, which often involves extensive prototyping and user feedback, does the team proceed to build and deliver the solution. This iterative, evidence-based approach stands in stark opposition to the output-focused project model, prioritizing learning and validation over mere execution of predefined features. The ability of product teams to generate 10-20 or more prototypes weekly has been a long-standing characteristic of this model, a capability further amplified by modern prototyping tools like Figma, even before the advent of generative AI.
A Shifting Bottleneck: The Impact of Technological Advancement
While these two models have coexisted for an extended period, the contemporary technological landscape has profoundly altered their dynamics. A pivotal change has been the dramatic reduction in the cost and complexity of product delivery. What once required significant investment in infrastructure, specialized talent, and lengthy development cycles can now be achieved with unprecedented speed and efficiency, thanks to advancements in cloud computing, open-source technologies, and, most recently, generative AI. This shift has fundamentally re-calibrated the bottlenecks in the product development lifecycle. Building a feature or an entire product is no longer the primary impediment; the challenge has moved upstream.
The project model, once considered a reliable if somewhat rigid framework, now often functions as a "turbo-charged feature factory." This accelerated production capability, paradoxically, can exacerbate the problem by enabling companies to create more "bad products"—products that fail to resonate with users, solve real problems, or achieve business objectives—at a faster rate than ever before. Industry analysis frequently cites high failure rates for new products, with some studies suggesting that upwards of 70-80% of new product launches do not achieve their intended market impact or profitability targets. This grim reality underscores the urgency of re-evaluating traditional development paradigms.
Consequently, the true bottleneck has unequivocally shifted from delivery to discovery. The critical challenge facing organizations today is not how to build, but what to build. This involves uncovering solutions that simultaneously address genuine customer needs and align with the company’s strategic goals, generating measurable and impactful outcomes. Furthermore, a successful solution must not merely solve a problem; it must solve it sufficiently better than existing alternatives, compelling customers to switch their allegiance. This competitive imperative intensifies the focus on unique value propositions and superior user experiences, pushing discovery to the forefront of strategic importance.
AI’s Dual Role: Enhancing Delivery and Reshaping Discovery
Generative AI plays a fascinating dual role in this evolving landscape. While it significantly aids in accelerating and optimizing delivery—for instance, through automated code generation, testing, and deployment—its contribution to discovery manifests in distinctly different ways. AI can assist in analyzing market data, identifying patterns in user feedback, and even generating initial conceptual ideas. However, it is in the nuanced and complex domain of product discovery that the product manager’s knowledge, intuition, and strategic insight remain absolutely critical. The human element of understanding unspoken needs, anticipating market shifts, and synthesizing diverse insights into a coherent product vision cannot yet be fully replicated by algorithms.
This paradigm shift presents a significant challenge for many product managers. As organizations transition from the project model to the more agile and outcome-focused product model, many product managers find themselves grappling with the definition of their future contribution. Their traditional responsibilities—project managing, facilitating meetings, documenting requirements—while valuable in the old framework, often fall short of providing the necessary strategic value in the new context. There is a palpable sense of uncertainty about how to adapt their roles to this new reality.
Some product managers, especially those with a strong technical background, have identified an opportunity to lean more directly into the building process. With the advent of sophisticated engineering tools like Claude Code and Cursor, which leverage generative AI to assist with coding, these individuals can actively participate in product construction, blurring the lines between product management and engineering. While this path is valid and beneficial for those possessing the requisite technical foundations, it is not the universal answer. The most effective product managers recognize that while they are indeed product builders and creators, their building serves a fundamentally different purpose than that of engineers.
"Build to Learn" vs. "Build to Earn": A Foundational Distinction
Years ago, renowned product coach and author Jeff Patton, celebrated for his seminal work "User Story Mapping: Discover the Whole Story; Build the Right Product," introduced the illuminating phrase "build to learn vs build to earn." This distinction precisely articulates the core difference between product discovery and product delivery. In the current era, particularly with the transformative capabilities of generative AI, Patton’s phrase resonates more powerfully than ever, providing a vital framework for product teams navigating the complexities of the modern product model.
Both discovery and delivery involve "building," but the purpose behind the building varies dramatically. In discovery, the primary objective is to build to learn; in delivery, it is to build to earn. This distinction also implies different tools, techniques, and, crucially, different interpretations of "testing."
Building to Learn: The Core of Product Discovery
In product discovery, the imperative is to build to learn. This phase is dedicated to exploring and validating a harmonious combination of technology, functionality, user experience, and business constraints that effectively addresses the four big risks inherent in any new product endeavor:
- Value Risk: Will customers actually use or buy this? Does it provide sufficient value?
- Usability Risk: Can users figure out how to use it? Is the experience intuitive and delightful?
- Feasibility Risk: Can we build this with our current technology and resources? Is it technically possible?
- Viability Risk: Does this solution work for our business? Can we support it, sell it, and make a profit?
The process of mitigating these risks relies heavily on rapid prototyping. Thanks to generative AI-based tools and platforms like Figma, creating 10-20 prototypes or prototype iterations per week is now achievable for virtually anyone on a product team, often without requiring direct engineering or dedicated design support for initial iterations. The fundamental purpose of these prototypes is to serve as instruments for testing hypotheses against these risks. "Testing" in this context involves:
- Value and Usability Testing: Engaging directly with target users and customers to gather qualitative and quantitative feedback on the prototype’s appeal and ease of use. This can range from informal interviews to sophisticated A/B tests on specific features.
- Feasibility Testing: Collaborating with engineers to assess the technical challenges, resource implications, and potential architectural complexities of the proposed solution. This ensures that what is being designed can actually be built robustly.
- Viability Testing: Presenting prototypes and discovery findings to internal stakeholders—including sales, marketing, legal, finance, and executive leadership—to ensure alignment with business objectives, regulatory compliance, and market strategy.
A significant innovation in this "build to learn" phase is the increased accessibility of live-data prototypes. Historically, creating functional prototypes that interact with real data was a costly and time-consuming endeavor. Today, the new generation of tools has drastically reduced this cost and time. This allows product teams to deploy functional, albeit limited, versions of their product to a select group of users and customers much earlier in the cycle. The ability to collect and analyze actual usage data from these live prototypes provides invaluable, unvarnished insights, accelerating the learning process and significantly de-risking the subsequent development phase. This represents a genuine game-changer for product discovery.
Furthermore, the speed and efficiency of modern prototyping tools enable product teams to test multiple approaches in parallel, rather than relying solely on sequential iteration. Traditionally, a team might identify what they believed was the optimal solution, iterate on it, and only proceed to full productization once sufficient evidence of success was accumulated. Now, it is increasingly common to simultaneously develop several distinct prototypes, each exploring a different facet or solution to the problem. These can then be tested concurrently, allowing for rapid comparison and identification of the most promising avenues, which can then be refined sequentially. This parallel experimentation dramatically reduces the time to uncover truly impactful solutions.
Building to Earn: The Rigors of Product Delivery
Once a solution has been thoroughly validated through the "build to learn" phase, the focus shifts to building to earn. This stage involves the development of a commercial-quality product—one that can be effectively sold, serviced, and supported, and upon which customers can reliably operate their businesses. The risks in this phase are fundamentally different and encompass a broad spectrum of non-functional requirements crucial for enterprise-grade software:
- Scale and Performance: Ensuring the product can handle anticipated user loads and maintain optimal performance under various conditions.
- Fault Tolerance and Reliability: Designing for resilience against failures, ensuring continuous availability and data integrity.
- Accuracy: Guaranteeing the precision and correctness of data processing and output.
- Privacy and Security: Implementing robust measures to protect user data and prevent unauthorized access, adhering to relevant regulations (e.g., GDPR, CCPA).
- Operations and Maintainability: Designing for ease of deployment, monitoring, and ongoing maintenance.
- Provisioning and Internationalization: Supporting diverse customer environments and adapting to global markets.
"Testing" in the delivery phase, therefore, takes on a different meaning. It involves a comprehensive suite of quality assurance processes to ensure the product meets this extensive list of demands and functions precisely "as advertised." This includes unit testing, integration testing, system testing, performance testing, security audits, and user acceptance testing (UAT), all geared towards guaranteeing a robust, reliable, and production-ready offering.
The Product Manager’s Evolving Skillset and Future Outlook
This profound shift in product development methodologies necessitates a re-evaluation of the skills and knowledge required for strong product management. While becoming proficient with prototyping tools and discovery techniques is increasingly straightforward, the more challenging and ultimately more valuable skill is the development of "product sense." Product sense is the intuitive ability to evaluate learnings from discovery, discern underlying user needs, anticipate market trends, and guide the product’s direction towards meaningful outcomes. It is a blend of empathy, strategic thinking, market insight, and creative problem-solving that is honed through experience and deliberate practice in the "build to learn" environment.
Leading companies have recognized this evolution and are actively adapting their product management interview processes to assess candidates’ understanding of this "build-to-learn" ethos and their proficiency in building and testing prototypes. They seek individuals who can not only articulate a vision but also demonstrate the practical ability to validate it through rapid experimentation. HR departments at major tech firms are reportedly revamping their recruitment frameworks to prioritize these capabilities, reflecting a broader industry trend.
However, not all individuals within the product community are equally enthusiastic about embracing this "build to learn" or "builder/creator" aspect of the role. Some product professionals prefer to view their contribution through the lens of facilitation, coordination, or simply being the "glue" that holds a team together. While these functions have their place, relying solely on them in the face of rapid technological change and the de-emphasis on delivery as a bottleneck places these individuals at increasing risk of obsolescence. The value proposition of a product manager in the modern era is intrinsically linked to their ability to drive discovery and validate solutions.
For those product managers who wholeheartedly embrace the builder/creator nature of their role, actively developing their product sense and mastering "build-to-learn" skills, a golden era of opportunity awaits. These individuals will be instrumental in navigating the complexities of product development in an AI-driven world, translating insights into impactful solutions, and shaping the future of innovation. Their ability to rapidly iterate, learn, and adapt will be a cornerstone of competitive advantage for organizations striving to deliver truly valuable products that stand out in an increasingly crowded marketplace. The emphasis on continuous learning and adaptability for product professionals is not merely a trend but a fundamental requirement for sustained success.
The shift from the project model to the product model, underpinned by the "build to learn vs. build to earn" philosophy, marks a maturation of the product development discipline. It signifies a move away from simply executing tasks towards strategically validating and delivering value. Organizations and product professionals who embrace this transformation are better positioned to thrive in a rapidly evolving technological landscape, ensuring that their efforts translate into genuine market impact and sustainable growth. The future of product management lies firmly in the realm of intelligent discovery and outcome-driven innovation.
