Key Takeaways
Treat software delivery as production, not a sequence of temporary projects
A new production age. Mik Kersten frames software as the defining means of production in a new technological era. Drawing on Carlota Perez’s model, he describes an Installation Period of rapid investment and startup growth, a Turning Point, and a Deployment Period when firms that master the technology scale it across the economy. The model is a lens for organizational change, not a precise timetable: the book argues that established businesses risk decline when they keep managing software with methods built for earlier eras.
Look at the factory. At BMW’s Leipzig plant, a car leaves the line about every seventy seconds, and the facility is designed around visible work and production constraints. Kersten contrasts this with enterprise IT, where leaders often cannot see how work moves or where it stalls. The lesson is not to copy an assembly line, but to connect business priorities to the work that produces software value.
The historical framing makes the management challenge feel structural rather than fashionable. It explains why simply buying new tools or renaming departments rarely changes an organization’s ability to deliver. The analogy also has limits: technological cycles are debated, and predictions about which firms will fail are uncertain. Its strongest use is therefore diagnostic, not prophetic. Leaders can ask whether their budgeting, authority, and visibility match software’s central role, without needing to believe that every industry follows a fixed fifty-year rhythm.
Measure Nokia's transformation by customer outcomes, not Agile ceremonies
Activity looked successful. Nokia tracked Agile adoption, team training, and tool use, yet developers described more urgent constraints: a slow build, test, and deployment loop, an architecture that resisted change, and a gap between team work and corporate planning. The transformation measured whether teams followed the process, not whether Nokia could deliver the software experience customers wanted. That mismatch helped hide the real bottlenecks as competitors built stronger mobile platforms.
The same trap scaled up. LargeBank, a major financial institution, launched a roughly $1 billion Agile and DevOps transformation measured largely by cost savings and project milestones. Digital product work ran separately from the IT changes needed to deliver it. Kersten’s point is not that Agile or DevOps caused either failure. Local improvements cannot rescue an end-to-end system if the true constraint sits elsewhere.
These cases make a useful distinction between adoption and capability. A firm can become better at performing a method while remaining unable to deliver a product customers choose. This resembles a common problem in public administration, where easily counted outputs can displace harder-to-measure outcomes. The corrective is not to discard process measures, but to treat them as diagnostic evidence. Outcome measures also need careful interpretation: weak product results may reflect market fit, distribution, or strategy, not only software delivery.
Fund enduring products and teams, then adjust investment as evidence changes
Projects expire; products persist. Project management assumes a defined scope, deadline, and end. Software products keep needing fixes, security work, new features, and maintenance after launch. Kersten argues that budgeting and staffing around temporary projects hides these continuing costs, encourages large up-front commitments, and repeatedly moves people away from the systems they know.
Keep teams with the work. Product-oriented management funds a product’s value stream, the activities required to deliver value to its users, and adjusts investment as demand and evidence change. Long-lived teams build knowledge and reduce the disruption of repeated reassignment. Boeing’s 787 illustrates the longer view: missing traceability between brake software and its requirements made future maintenance and certification harder, so rewriting working software could be economically preferable over a decades-long support life. Product funding does not eliminate planning; it makes planning recurrent.
The shift changes the unit of management from a temporary undertaking to a continuing economic capability. That fits software’s ongoing evolution and resembles portfolio approaches in research and product design, where funding can be renewed as uncertainty resolves. Stable teams also preserve tacit knowledge that formal documentation rarely captures. Still, permanence is not automatically virtuous. Products should have explicit review and retirement decisions, or product funding can protect low-value systems just as project budgets can conceal them.
Track four kinds of work: features, defects, risks, and technical debt
Name what moves. A flow item is a unit of work that delivers value through a product’s value stream. Kersten groups software work into four categories: features add customer-facing value; defects repair quality problems; risks address security, privacy, governance, or compliance; and debts remove obstacles that could make future delivery harder. The categories help business and technical teams discuss work in a shared vocabulary.
Make hidden work visible. Security controls may be invisible to users until they fail, while technical debt can remain unnoticed until changes become slow or costly. The framework treats both as real work that competes with feature delivery. Its advice is to classify work by the value it supports, rather than treating every architecture improvement as automatically worthwhile. Debt reduction earns priority when it enables future delivery or reduces meaningful operational risk.
This taxonomy is valuable because organizations often count visible features while leaving risk and maintainability in vague side channels. Naming them makes tradeoffs discussable before a breach or a stalled release forces attention. The categories also prevent a false choice between innovation and upkeep: both can contribute to durable product value. A caution is that classification is not valuation. Teams still need sound product, security, and engineering judgment to decide which risk or debt matters most.
Use flow metrics to find where work waits, accumulates, or slows
Five views of flow. Flow metrics describe how work moves through a value stream. Flow distribution shows the share of work devoted to each flow item. Flow velocity counts items completed over time. Flow time measures elapsed time from accepting work to delivering it. Flow load counts active or waiting work, a form of work in progress. Flow efficiency compares time actively spent working with total elapsed time, revealing delays and queues.
Choose targets by product. A new product may need more feature work; a mature service may need more defect or risk work. Kersten describes a product launch where feature demands and unexpectedly fast customer adoption crowded out debt reduction. The resulting slowdown in future feature work showed why teams need to forecast capacity and make tradeoffs explicit. These metrics are most useful for following a particular value stream over time, not ranking unlike products.
The set is useful because it separates amount, speed, workload, and waiting, which are often collapsed into a single dashboard number. In queueing systems, adding work to an already crowded pipeline can increase delay rather than output, so load and wait time deserve attention alongside completion counts. Flow velocity, however, counts items rather than their size or business impact. Comparisons can mislead when teams define work differently, so trends and context matter more than a league table.
Connect delivery metrics to value, cost, quality, and team happiness
Measure the business result. The Flow Framework pairs delivery data with four business outcomes for each product value stream: value, cost, quality, and happiness. Value might be revenue, recurring revenue, or adoption for an internal product. Cost includes the people, infrastructure, and other delivery resources supporting it. Quality can include customer-visible defects, incidents, renewals, or satisfaction. Happiness tracks staff engagement, including measures such as employee Net Promoter Score.
Put outcomes beside work. A dashboard can show whether more feature delivery coincides with revenue growth, whether defects increase after a rushed release, or whether a team’s engagement falls as manual work accumulates. Internal platforms can be assessed through adoption or the value they enable in dependent products. The framework asks organizations to define measures suited to their context, rather than pretending one universal revenue number fits every product.
The four-part view is stronger than cost-only governance because it makes tradeoffs visible: reducing expense while damaging quality or team health is not an unambiguous success. It also encourages leaders to examine whether software activity produces a business result at all. Attribution remains difficult, however. Revenue depends on pricing, sales, and market conditions, while employee surveys can be distorted by fear or survey fatigue. Metrics should prompt investigation and conversation, not substitute for them.
Build a connected value-stream network across specialized tools
Specialization creates seams. A Value Stream Network is the connected system of people, activities, processes, tools, and work artifacts through which software value moves. In Kersten’s study of 308 enterprise tool networks, organizations used 55 different tools; 69.3 percent had work artifacts spanning three or more tools. Different specialists need different systems, so forcing every role into one tool may sacrifice useful capabilities.
Connect instead of flattening. The book distinguishes fundamental complexity, which supports real differences in roles or scale, from accidental complexity, such as redundant tools with no value-adding purpose. Integrating tools can reduce duplicate entry and lost handoffs while allowing specialists to keep the systems suited to their work. The goal is not tool uniformity. It is reliable flow and visibility across the tools an organization actually uses.
This is a more realistic answer to organizational complexity than either demanding one universal platform or tolerating disconnected silos. The analogy to medical specialization is apt: expertise deepens when roles specialize, but coordination failures can undermine the benefit. Integration itself is not free, and poorly governed links can reproduce bad data at greater speed. Organizations need to decide which differences genuinely serve work, then make the remaining handoffs explicit and maintainable.
Route software work through collaboration networks, not imaginary assembly lines
Software work takes alternate routes. Physical manufacturing often follows a repeatable sequence. Software delivery combines design and production, handles changing requirements, and depends on people who can reroute work when a team or component is blocked. Kersten’s third epiphany is that software value streams resemble networks more than straight production lines. A dependency may slow a team, but another team can sometimes work around it rather than stopping the whole system.
Use three connecting models. The Integration Model maps artifacts across tools to a common structure. The Activity Model maps detailed workflow states to shared activities and flow states. The Product Model connects those work records to the products and value streams the business manages. Together, they turn scattered tool data into a view of work that can be related to business outcomes.
The network model captures a distinctive feature of knowledge work: adaptation is part of production, not an exception to it. That makes the metaphor more than a visual choice, since a linear model can lead leaders to optimize one stage while ignoring alternate routes and dependencies. Networks can also be harder to govern and interpret. Their usefulness depends on accurate relationships and clear product boundaries, not on drawing a complicated diagram and calling it visibility.
Give security and technical debt explicit capacity before platform failure
Invisible work has consequences. The book connects Equifax’s 2017 breach, which exposed data from about 145.5 million consumer accounts, to the danger of treating security as an individual developer’s mistake rather than a business-wide responsibility. It also links Nokia’s decline to platform debt that made adapting its software increasingly difficult. In automotive software, the book cites 14 million recalls in 2016 related to software issues, showing how feature pressure can make quality costly.
Make the tradeoff deliberate. Microsoft offers a contrasting example: Bill Gates shifted company-wide attention first toward internet features, then toward security through Trustworthy Computing. In the Flow Framework, this means deliberately changing the share of work allocated to features, defects, risks, and debts. The cases suggest that leaders need visibility to make such choices; they do not prove that one metric or allocation alone would have prevented every failure.
The strongest lesson is systemic accountability. Security, reliability, and maintainability cannot be delegated to a lone programmer if the organization’s priorities and incentives determine what receives attention. Making capacity explicit gives boards and executives a chance to discuss risk before it becomes a headline. Yet metrics cannot replace technical controls, threat modeling, or incident learning. The cases are persuasive illustrations of the framework, not controlled evidence that a different dashboard would necessarily have changed each company’s fate.
Analysis
The book’s central contribution is a change in managerial unit. It argues that software should be governed as a continuing portfolio of products and value streams, rather than as temporary projects whose success is judged by schedule and cost. Its examples make that argument concrete: Nokia could measure Agile activity without improving the mobile platform; LargeBank could spend heavily on transformation while separating digital ambitions from delivery capacity; BMW could make production constraints visible because business needs and work were closely connected.
The Flow Framework’s strength is its translation layer. Four kinds of work and a small family of flow measures give technical and business leaders a vocabulary for discussing where investment goes and what it produces. The proposed network architecture addresses an equally practical obstacle: enterprise work lives across specialized tools, so meaningful end-to-end measurement requires connected artifacts and clear product boundaries.
The framework should be treated as a decision aid, not an automatic optimization engine. Counting completed items does not establish their relative value, and identical metrics may mean different things across products. Outcome attribution is also difficult: market response depends on sales, pricing, competition, and customer needs as well as delivery. Finally, measuring employee happiness can help reveal friction, but a survey cannot capture trust or replace direct conversation.
A careful implementation would use the metrics to ask better questions, not to reward whichever team produces the cleanest numbers. Start with a clearly bounded product, map its actual work, establish baseline measures, and examine unexpected changes with the people doing the work. Preserve team autonomy while making tradeoffs visible to leadership. The framework is most compelling when it helps an organization learn where value is blocked, then change investment or structure in response. Its promise is not certainty, but a more informed way to manage software as a core business capability.
Review Summary
Project to Product receives mixed reviews, with an average rating of 3.93/5. Many praise its insights on transitioning from project to product-based development and the Flow Framework for managing value streams. Readers appreciate the book's emphasis on business value and alignment with digital transformation. However, some criticize it for lacking practical implementation details and being overly focused on large enterprises. The BMW manufacturing analogy is contentious, with some finding it helpful and others seeing it as misaligned with software development realities.
People Also Read
Glossary
Installation Period
early technology adoptionIn Carlota Perez’s model of technological revolutions, the Installation Period is the early phase when a new technology attracts investment, infrastructure, and new businesses. These ventures challenge industries built around the previous era’s production methods. The period leads into a Turning Point and a later Deployment Period, when the new means of production becomes more widely established.
Deployment Period
scaled use of technologyIn Perez’s model, the Deployment Period follows the Turning Point of a technological revolution. Organizations that have mastered the new means of production use it at scale, and the associated infrastructure spreads through the economy. In this book, the term describes the anticipated phase in which software becomes a central production capability across established businesses.
Flow Framework
business-aligned delivery managementA management framework for connecting software delivery to business results. It organizes work around product value streams, classifies work into four flow items, and relates delivery measures to outcomes such as value, cost, quality, and team happiness. It is intended to complement Agile and DevOps practices by giving business and technical leaders a shared view of end-to-end delivery.
value stream
end-to-end value deliveryThe complete set of activities, people, processes, and tools involved in delivering value to a customer through a product or service. A software value stream can span business planning, design, development, testing, operations, and support. The customer may be external or an internal user of a platform or application.
flow item
unit of delivered workA unit of work that moves through a product’s value stream and represents business value or a necessary investment. The framework groups flow items into features, defects, risks, and debts. These categories make different kinds of software work visible and help teams discuss their competing demands.
flow distribution
allocation across work typesThe proportion of a value stream’s work devoted to each flow item category over a period of time. A product preparing for launch may allocate more capacity to features, while a service facing security exposure may prioritize risks. The distribution makes the tradeoffs among work types explicit; increasing one share limits capacity available for others.
Value Stream Metrics
measures of delivery flowA group of measures used to understand how work moves through a software value stream. Flow distribution tracks the mix of work, flow velocity counts completed items over time, flow time measures elapsed delivery time, flow load tracks active or waiting work, and flow efficiency compares active work time with total elapsed time.
Value Stream Network
connected delivery systemThe connected system of teams, activities, tools, and work artifacts involved in delivering software value. Unlike a simple linear pipeline, it can contain dependencies, branches, and alternate routes for work. The framework uses the network to make end-to-end delivery visible and connect work to products and business outcomes.
Integration Model
common model across toolsThe model that maps work artifacts and related information across the different tools in a delivery system. It gives tools with different schemas a shared structure, so work can be connected across repositories and mapped to the framework’s flow items. It also allows tools to be added or replaced without rebuilding every other integration.
Activity Model
shared workflow interpretationThe model that maps detailed workflow states used in tools to common activities and the framework’s generic flow states. Organizations may have many local states for work such as review, testing, or approval. The Activity Model provides a consistent basis for measuring work across those different workflows.
Product Model
work mapped to productsThe model that connects artifacts and work containers in the tool network to product-oriented value streams. It lets an organization relate delivery activity and its associated costs and results to the products that customers or internal users consume. The model also helps reveal work that has not been aligned to a product.
Connectivity Index
coverage of tool integrationsA measure of how much of the tool network has been connected through integrations. It compares connected repositories and artifact containers with those that remain unconnected. A low index warns that flow measures may omit portions of the delivery process, such as customer support or deployment.
Traceability Index
coverage of artifact linksA measure of how fully work artifacts are connected to related artifacts across a value stream. Traceability can link a customer request to requirements, code changes, tests, and releases. A higher index gives a more complete view of work and can support reporting, maintenance, governance, and compliance.
Alignment Index
coverage of product mappingA measure of how much work represented in the tool network has been mapped to a product value stream. It compares artifact containers connected to products with the total set of containers. A low index indicates that some organizational work remains outside the product-oriented view.
FAQ
What's Project to Product about?
- Digital Transformation Focus: Project to Product by Mik Kersten explores how organizations can transition from a project-oriented approach to a product-oriented mindset in the Age of Software. It emphasizes adapting to digital disruption.
- Flow Framework Introduction: The book introduces the Flow Framework, a managerial tool that connects technology investments to business outcomes, aiming to improve efficiency and value delivery.
- Real-World Examples: Kersten uses case studies from companies like BMW and Nokia to illustrate the challenges and successes of digital transformation, highlighting the pitfalls of traditional project management.
Why should I read Project to Product?
- Relevance to Modern Businesses: Essential for those involved in software delivery, product management, or digital transformation, as it addresses the urgent need to adapt to the Age of Software.
- Practical Framework: Offers actionable insights and a practical framework (the Flow Framework) to improve software delivery and align it with business goals.
- Expert Insights: Mik Kersten provides a unique perspective on the intersection of technology and business, helping organizations avoid common pitfalls in their transformation journeys.
What are the key takeaways of Project to Product?
- Shift to Product Management: Organizations must transition from project-based thinking to a product-oriented mindset to remain competitive and deliver continuous value.
- Importance of Value Streams: Understanding and managing value streams is crucial for optimizing software delivery and ensuring customer satisfaction.
- Integration of Business and IT: Better integration between business strategies and IT operations is necessary, with the Flow Framework providing a common language for alignment.
What is the Flow Framework in Project to Product?
- New Management Approach: The Flow Framework helps organizations visualize and manage their software delivery processes, connecting technology investments to business outcomes.
- Four Flow Items: It identifies four key flow items—features, defects, risks, and debts—that represent different types of work in a software value stream.
- End-to-End Visibility: Provides end-to-end visibility of business value flow, allowing organizations to identify bottlenecks and optimize processes.
How does Project to Product address digital disruption?
- Understanding Digital Disruption: Explains that digital disruption is a predictable outcome for companies that fail to adapt to new production means, especially in the Age of Software.
- Case Studies: Uses real-world examples, such as Nokia's failure to adapt, to illustrate the consequences of ignoring digital disruption.
- Framework for Adaptation: The Flow Framework offers a structured approach for aligning software delivery processes with business goals to maintain competitiveness.
How does the Flow Framework help in measuring success?
- Flow Metrics: Provides specific metrics, such as flow time and flow velocity, to measure performance in delivering value and identify areas for improvement.
- Business Correlation: Correlates flow metrics with business results, helping organizations understand the impact of software delivery on overall performance.
- Continuous Improvement: Encourages a culture of continuous improvement by providing insights into bottlenecks and inefficiencies.
What are the four flow items defined in Project to Product?
- Features: Represent new business value delivered to customers, prioritized based on customer needs.
- Defects: Issues affecting product quality that must be addressed to maintain customer satisfaction.
- Risks: Encompass security, compliance, and governance issues that need proactive management.
- Debts: Refer to technical debt that can hinder future development efforts, requiring prioritization for reduction.
What are the Management Zones mentioned in Project to Product?
- Three Management Zones: The Performance Zone, Transformation Zone, and Incubation Zone represent different stages of product development.
- Performance Zone: Focuses on delivering high-quality, stable products that generate revenue, prioritizing efficiency and reliability.
- Transformation and Incubation Zones: The Transformation Zone is for innovation and adaptation, while the Incubation Zone explores new ideas and technologies.
How does Project to Product suggest measuring success?
- Value Stream Metrics: Emphasizes measuring success through value stream metrics that track the flow of business value in software delivery.
- Business Outcomes: Success should be measured in terms of business outcomes like revenue, customer satisfaction, and employee engagement.
- Continuous Improvement: Encourages a mindset of continuous improvement, using metrics to identify areas for enhancement.
How does Project to Product address the concept of technical debt?
- Definition of Technical Debt: Described as the cost of future rework due to shortcuts taken in the present, impacting software delivery.
- Impact on Flow: Excessive technical debt can slow down flow velocity and increase flow time, affecting value delivery.
- Strategies for Management: Provides strategies for addressing technical debt, such as allocating time for debt reduction and prioritizing it alongside feature development.
What role do case studies play in Project to Product?
- Real-World Examples: Uses case studies from companies like BMW and Nokia to illustrate concepts and frameworks, providing context for readers.
- Lessons Learned: Highlights key lessons from successes and failures in digital transformation, helping readers avoid common pitfalls.
- Illustrating the Flow Framework: Demonstrates how the Flow Framework can be applied in real-world scenarios, showcasing its effectiveness.
What are the best quotes from Project to Product and what do they mean?
- "Organizing software development as a group of loosely connected projects will never lead to good products.": Emphasizes the importance of a product-oriented approach over a project-based mindset.
- "The Flow Framework is a new way of seeing and measuring delivery and aligning all of your IT investments according to value streams.": Highlights the framework's role in transforming software delivery management.
- "If you want to get rid of obsolete practices and succeed with the new digital, read this book.": Encourages organizations to embrace new methodologies to remain competitive.
Download PDF
Download EPUB
.epub digital book format is ideal for reading ebooks on phones, tablets, and e-readers.