The Enduring Dominance of the Project Model and Its Limitations
Historically, and remarkably still prevalent even amidst the AI revolution, the project model has been the default approach to bringing new features and products to market. This methodology is fundamentally centered around output. In this framework, the impetus for development typically originates from executive mandates or stakeholder roadmaps, which outline a prioritized list of features or projects. Once these are defined, the process becomes largely sequential: a product manager, often functioning as a "feature team product manager," drafts a detailed specification, such as a Product Requirements Document (PRD). This document then guides designers in creating user interfaces and experiences, which are subsequently handed over to engineers for implementation. The success metric here is often the timely delivery of the specified features, regardless of their actual impact or adoption.
This output-driven approach, while offering a semblance of control and predictability in planning, has long been criticized for its inherent inefficiencies and propensity to generate what industry experts term a "feature factory." A feature factory, by its nature, prioritizes the volume and speed of feature production over their strategic value or customer utility. Data from various industry reports consistently highlights the shortcomings of this model; for instance, studies by organizations like the Standish Group have historically indicated high failure rates for IT projects, with many failing to meet original objectives, running over budget, or being abandoned entirely. More specifically, product-focused research often reveals that a significant percentage of features—some estimates suggest as high as 60-80%—are rarely or never used by customers. This stark reality underscores the critical flaw of the project model: its disconnect from genuine customer problems and business outcomes. The emphasis on delivering a predefined scope, rather than discovering a valuable solution, often leads to a proliferation of bad products, delivered faster than ever before, thanks to modern engineering efficiencies.
The Emergence of the Outcome-Driven Product Model
In stark contrast to the output-centric project model, the product model is unequivocally focused on outcomes. This paradigm begins not with a predetermined list of features, but with the identification of a significant problem that needs solving for either the customer or the business. At the heart of this model lies the empowered, cross-functional product team. These teams, comprising product managers, designers, and engineers, are granted the autonomy and responsibility to not only identify problems but also to discover and validate solutions. Their primary mandate is to deliver measurable outcomes, such as increased customer retention, higher conversion rates, or improved operational efficiency, rather than merely shipping features.
A hallmark of the product model is its iterative and experimental nature, particularly during the discovery phase. It is not uncommon for product teams operating within this framework to generate and test numerous prototypes—often 10 to 20 or more per week. This rapid prototyping cycle, a practice that predates the recent surge in generative AI, was already facilitated by advanced design tools like Figma, which allowed for quick iteration and visualization. This iterative discovery process ensures that solutions are validated with real users and stakeholders, backed by evidence, before significant development resources are committed. This approach significantly de-risks product investments by ensuring that what is built genuinely addresses a need and contributes to desired outcomes.
The Shifting Bottleneck: From Delivery to Discovery
A pivotal development underpinning the ascendancy of the product model is the dramatic reduction in the cost and complexity of product delivery. Over the past decade, advancements in cloud computing, DevOps practices, open-source software, and low-code/no-code platforms have revolutionized the engineering landscape. What once required significant infrastructure investment and lengthy development cycles can now be deployed with unprecedented speed and efficiency. The advent of generative AI tools, such as Claude Code and Cursor, further accelerates this trend by automating significant portions of the coding process, making the actual construction of features faster and cheaper than ever before. This technological progress has effectively shifted the bottleneck in product development.
Where once the challenge lay in the sheer effort and time required to build a product, the new bottleneck resides firmly in discovery: identifying a solution that is genuinely worth building. A truly valuable solution is one that simultaneously addresses a critical customer pain point and delivers tangible benefits to the company. It must generate the necessary outcome, and crucially, it must be demonstrably superior to existing alternatives, compelling customers to switch or adopt it. This emphasis on superior value proposition necessitates deep customer understanding, rigorous experimentation, and a keen sense of market dynamics—areas where the product manager’s knowledge and expertise become indispensable.
AI’s Dual Impact: Enhancing Delivery and Revolutionizing Discovery
Generative AI’s influence extends across both product delivery and discovery, albeit in profoundly different ways. In delivery, AI tools are streamlining the engineering process by assisting with code generation, automated testing, debugging, and even deployment orchestration. This empowers engineering teams to execute on well-defined solutions with unparalleled speed and accuracy.
However, its impact on discovery is arguably more transformative for the product management discipline. While AI cannot replace human creativity, empathy, or strategic judgment, it significantly augments the discovery process. AI can analyze vast datasets to uncover user patterns, identify unmet needs, and predict market trends. It can accelerate market research, generate a multitude of design concepts, simulate user interactions, and even provide rapid feedback on prototype iterations. This capability allows product teams to explore a wider range of potential solutions and validate hypotheses with greater speed and precision, moving beyond intuition to data-backed decisions. Yet, it is precisely in the nuanced evaluation of these AI-generated insights and the strategic guidance of the discovery process that the product manager’s unique blend of domain knowledge, customer empathy, and business acumen becomes irreplaceable.
The Product Manager’s Evolving Contribution: "Build to Learn" vs. "Build to Earn"
The paradigm shift from the project to the product model, amplified by AI, presents a significant challenge and opportunity for product managers. Many seasoned product managers, accustomed to their traditional roles of project coordination, roadmap management, and specification writing, find themselves at a crossroads. Their previous contributions, often centered around facilitation and process management, no longer provide the requisite value in an outcome-driven, discovery-intensive environment.
Some product managers with strong technical foundations are leveraging the new generation of engineering tools (e.g., Claude Code, Cursor) to directly contribute to the building of the product, blurring the lines between product and engineering roles. While this path is valid for those with the appropriate skill set, the most impactful product managers are realizing a more profound truth: they are indeed product builders and creators, but their purpose in building differs fundamentally from that of engineers.
This distinction was eloquently captured by product coach Jeff Patton, author of "User Story Mapping: Discover the Whole Story; Build the Right Product," who coined the phrase "build to learn vs. build to earn." This conceptual framework resonates particularly powerfully with product teams navigating the age of generative AI.
In product discovery, we are building to learn. The objective is not to create a polished, shippable product, but to rapidly test hypotheses and mitigate key risks associated with a potential solution. These risks, often categorized as value (will users want it?), usability (can users use it?), feasibility (can we build it?), and viability (should we build it from a business perspective?), are systematically addressed through experimentation. The rapid prototyping capabilities afforded by modern tools, including AI-powered design platforms, enable product managers to create 10-20 (or more) prototype iterations per week independently, without necessarily relying on designers or engineers for every iteration. "Testing" in this context involves engaging users and customers to validate value and usability, consulting engineers for feasibility assessments, and collaborating with stakeholders to ensure business viability. Live-data prototypes, which allow functional prototypes to be placed in front of select users to collect real usage data earlier and more affordably, represent a game-changer for this "build to learn" phase. Furthermore, the speed of generative AI tools allows for parallel testing of multiple approaches, accelerating the identification of the most promising solutions.
Conversely, in product delivery, we are building to earn. This phase focuses on constructing a commercial-quality product that is robust, scalable, and ready for market. The risks here are entirely different, encompassing concerns such as scale, performance, fault tolerance, reliability, accuracy, privacy, security, operational efficiency, provisioning, and internationalization. "Testing" in delivery means rigorously ensuring the product meets this exhaustive list of demands and functions precisely as advertised, ready for customers to integrate into their businesses.
Implications for the Future of Product Management
This emphasis on "build to learn" has profound implications for the skills and knowledge required of contemporary product managers. While proficiency with prototyping tools and discovery techniques is increasingly table stakes, the true differentiator lies in developing "product sense." Product sense is the intuitive ability to evaluate learning from experiments, synthesize diverse data points, understand market dynamics, anticipate user needs, and ultimately guide the product’s direction with strategic foresight. It combines deep empathy for the customer, a nuanced understanding of technology’s capabilities, and sharp business acumen.
Leading companies are actively adapting their product management interview processes to assess candidates’ understanding of this builder/creator nature of the role. Interviews now frequently include challenges that require candidates to articulate discovery strategies, propose prototyping approaches, and demonstrate their ability to iterate based on simulated user feedback. This shift underscores a broader industry movement towards continuous discovery and product-led growth, where experimentation and validated learning are paramount.
However, this transition is not universally embraced. Some product managers, who prefer roles centered on facilitation, coordination, or acting as "glue" for teams, may find their traditional value proposition diminishing. As automation and AI streamline many administrative and project management tasks, these roles are increasingly at risk of obsolescence or significant reduction in scope.
For those product managers who wholeheartedly embrace the builder/creator ethos, focus on cultivating their product sense, and become adept at "build-to-learn" methodologies, the future appears exceptionally bright. We are entering a golden era for skilled product professionals, where their ability to rapidly discover, validate, and guide the creation of truly valuable products will be a critical competitive advantage for organizations across all sectors. This evolving landscape demands not just managers of products, but architects of outcomes, capable of navigating complexity with agility, insight, and a relentless commitment to learning. The product manager is no longer merely an orchestrator but a primary driver of innovation, standing at the forefront of value creation in the digital age.