Professional

From Output to Outcome: Why Shipping Features Is Not the Same as Delivering Value

Automatic translation · View original

Topic illustration: From Output to Outcome: Why Shipping Features Is Not the Same as Delivering Value

Some teams ship. And ship. And ship. Sprint after sprint, the backlog moves forward, demos showcase new features, and stakeholders applaud. But when someone takes a closer look at the business numbers (retention, conversion, real user satisfaction), the picture is far less bright.

We explore why so many agile teams still confuse motion with progress, and what that confusion means for those of us working in product.

The Mirage of Velocity


There is a very well-camouflaged trap at the heart of modern agile development: velocity as a proxy for value. Teams measure story points completed, features launched, epics closed. Product managers build roadmaps full of specific deliverables. Stakeholders approve quarters full of items that "are going to be done." And so, without anyone explicitly deciding it, the organization ends up measuring its capacity to produce things instead of its capacity to generate real change in the world.

This distinction has a name in the product literature. We speak of the difference between output—what the team produces—and outcome—the change in behavior or condition that output brings about in users or in the business. And although the distinction seems obvious when stated this way, in practice it is extraordinarily difficult to sustain.

Josh Seiden (2019) explains it with disconcerting clarity: organizations are structured to produce things, not to generate change. Their planning processes, their tracking metrics, their reporting dynamics and their conversations with leadership revolve around the question "what are you going to deliver?" instead of "what is going to change?" This is not irresponsibility. It is systemic inertia.

How many times have we taken part in a quarterly planning session where the central conversation was which features get in and which don't? How often did someone in that room ask the reverse question: what user behavior do we need to change, and how will we know we have achieved it?

Why the Confusion Persists


It would be unfair to blame teams alone for this dysfunction. The pressure toward output has deep roots that deserve to be named.

First, outputs are tangible and outcomes are not. A feature can be shown. You can click on it. You can photograph it in a demo. A change in user behavior, by contrast, requires time, metrics, and a prior hypothesis in order to be measured. The intrinsically more observable nature of output makes it the perfect currency in organizations that need to justify investment in the short term.

Second, uncertainty is uncomfortable. Working oriented toward outcomes means accepting that we do not know with certainty which feature will produce the desired change. It means saying "this is our hypothesis" instead of "this is what we are going to build." For organizational cultures accustomed to predictability as a signal of competence—something Amy Edmondson has studied extensively in her research on psychological safety and learning cultures—admitting uncertainty can feel like weakness, even though it is exactly the opposite.

Third, incentive structures reinforce output. Teresa Torres (2021) points out that most product teams are evaluated by how much they build, not by how much they learn or what results they generate. If a Product Manager is judged by the number of features launched in the quarter, they have every incentive in the world to optimize that metric, even if it connects with no real strategic objective.

Marty Cagan develops this argument more explicitly: the difference between product teams that generate real impact and those that simply execute roadmaps lies precisely in whether they have the autonomy and responsibility to pursue outcomes or whether they are basically feature factories with agile methodology as decoration.

The Feature Roadmap Trap


One of the most revealing symptoms of the output culture is the feature roadmap. That document (tool, board, or presentation) is full of specific items organized over time: "In Q1, advanced search filters. In Q2, CRM integration. In Q3, dashboard redesign."

The problem is not that these items are wrong. The problem is that their presence on the roadmap precedes the conversation about which user or business problem they are solving, and what evidence exists that this specific solution is the right one. As Melissa Perri argues, many organizations fall into what she calls the "build trap": teams so busy delivering that they never stop to ask whether what they deliver matters.

What alternative exists? Perri, Torres, Seiden and other authors in the field converge on one direction: problem-oriented or outcome-oriented roadmaps. Instead of committing to specific solutions over time, the team commits to the problem it will address and to the indicators that will tell it whether it is solving it. How exactly it will solve it is a hypothesis that will be refined with evidence.

This does not mean product roadmaps cannot have any specificity. It means specificity should come after validation, not before. And at the strategic level, what should guide planning are questions like: "What needs to change in our users' behavior for our business to grow?" or "What friction is preventing users from getting what they came for?"

Outcomes, Metrics, and False Precision


Adopting an outcome orientation does not automatically solve the problem. It generates a new one: choosing the right outcomes and measuring them without falling into a different but equally harmful trap—that of proxy metrics that end up being confused with the real objective. This phenomenon has a name: Goodhart's Law, originally formulated by economist Charles Goodhart in 1975 and popularized in management contexts: when a measure becomes a target, it ceases to be a good measure. A team that sets as its outcome "increase the number of users who complete onboarding" may achieve it by simplifying the process to the point where users who complete onboarding do not really understand the product. The metric improves. The business does not.

That is why OKRs, for example, can be useful, but they require discipline in their application. Key Results should measure observable changes in behavior or condition, not activities. "Launch three new integrations" is an output disguised as a KR. "Increase by 15% the percentage of users who use the product more than three times a week" is a measurable outcome.

The distinction made by Jeff Gothelf and Josh Seiden (2021) between behavior metrics and activity metrics is very useful here. Activity metrics count things the team does (features launched, sprints completed, tickets closed). Behavior metrics count things that change in users or in the business as a result of what the team does. Only the latter are outcomes.

That said, not every organization has the analytical maturity to measure user behavior with the granularity that some ideal frameworks presuppose. And here it is worth being honest: an outcome orientation is not a lever that can be switched on overnight. It requires product instrumentation, a culture of experimentation, and above all, time. For many teams, the path toward outcomes goes through intermediate stages that should not be underestimated.

The Role of Discovery and Continuous Validation


If outcomes are the north star, continuous discovery is the process that allows us to navigate toward it. Teresa Torres has rigorously articulated how high-performing product teams integrate user research as a continuous activity, not as a phase that precedes development.

In Continuous Discovery Habits, Torres proposes the Opportunity Solution Tree as a tool to connect the desired outcome with the opportunities identified in user research, and these with the possible solutions to explore. This visual structure is not just a planning artifact—it is a way of thinking that keeps the causal chain between what is built and the change it is meant to generate visible.

The risk of omitting this causal chain is building brilliant solutions to problems users do not have, or to versions of the problem that the team imagined without sufficient validation. This pattern is not exclusive to startups—it is reproduced in internal teams of large organizations that, with all the resources in the world, build features nobody uses.

How do you know, in your team, how many of the features delivered in the last year have evidence of sustained use? How many were launched, generated some initial activity, and then fell into oblivion? These are not rhetorical questions: they are the kind of audit that mature teams perform regularly.

The Organizational Dimension: When the Problem Is Not in the Team


At this point, it is worth resisting the temptation to reduce this problem to a matter of methodology or individual skills. An outcome orientation requires certain organizational conditions that are not always present, and it would be dishonest to ignore this. Cagan is categorical on this point: teams cannot pursue outcomes if the organization dictates outputs. If leadership presents roadmaps committed to specific features to clients or investors, if contracts with suppliers specify deliverables, if the performance culture evaluates the team by what it produces and not by the impact it generates—in that context, asking the team to "think in outcomes" is asking it to swim against the current of the entire system in which it operates.

This does not mean giving in. It means being strategic. Some changes can start at the team level: formulating explicit hypotheses before building, measuring the impact of what has been delivered, proposing success metrics before planning. Other changes require conversations with leadership about what the organization really wants to measure. And some, to be frank, are not possible without a deeper transformation that goes beyond what a product team can manage alone.

Dave Snowden and his Cynefin framework remind us that in complex environments there are no universal best practices, only good practices for a specific context. The question is not "should my organization orient itself toward outcomes?" but rather "what is the path toward that orientation given my specific organizational context?"

Small Steps Within Our Reach


Without claiming to offer universal recipes, there are some practices that different teams have found useful for moving toward a greater outcome orientation, regardless of the organization's maturity level.

  1. Formulate hypotheses before committing to solutions. Before writing a user story, ask yourself:
  • What user behavior do we want to change with this?
  • How will we measure it?
  • What evidence do we have that this solution is the right one?

This habit, practiced consistently, changes the nature of planning conversations.

  1. Include success criteria in user stories. Not only "as a user I want to be able to X" but also "we will know this feature is successful when we see Y." This forces the team to think about impact before building, and provides a basis for evaluating decisions after launch.
  2. Review the impact of what has been delivered. Regularly dedicate time—it can be part of the retrospective, or a separate session—to reviewing usage and behavior metrics for features launched in previous cycles. What worked as we expected? What did not? What did we learn?
  3. Propose outcome metrics before planning. Instead of arriving at planning with a list of features, arrive with a question: "Which indicator do we want to move this quarter, and by how much?" This redirects the conversation toward impact before committing to solutions.

None of these practices is a silver bullet. They are entry points. What makes them work (or not) depends largely on the organizational context and on the willingness of the people with authority to sustain them over time.


In the end, the distinction between output and outcome is not just a methodological discussion. It is a question about the purpose of the work we do. Are we here to build things or to generate change? Do we measure our capacity to produce or our capacity to learn and adapt?

There is no single answer. But the way an organization responds to these questions largely defines what kind of product team it is capable of being.

Where would you place your team today on this spectrum? What conversation would you need to have with your organization for the outcome orientation to be something more than a declared ideal?

Bibliography


  • Cagan, M. (2017). Inspired: How to Create Tech Products Customers Love (2nd ed.). Wiley.
  • Cagan, M., and Jones, C. (2020). Empowered: Ordinary People, Extraordinary Products. Wiley.
  • Doerr, J. (2018). Measure What Matters: OKRs — The Simple Idea That Drives 10x Growth. Portfolio/Penguin.
  • Edmondson, A. C. (2018). The Fearless Organization: Creating Psychological Safety in the Workplace for Learning, Innovation, and Growth. Wiley.
  • Gothelf, J., and Seiden, J. (2021). Lean UX: Designing Great Products with Agile Teams (3rd ed.). O'Reilly Media.
  • Goodhart, C. A. E. (1975). Problems of Monetary Management: The U.K. Experience. Papers in Monetary Economics, Reserve Bank of Australia.
  • Perri, M. (2018). Escaping the Build Trap: How Effective Product Management Creates Real Value. O'Reilly Media.
  • Seiden, J. (2019). Outcomes Over Output: Why Customer Behavior Is the Key Metric for Business Success. Sense & Respond Press.
  • Snowden, D. J., and Boone, M. E. (2007). A Leader's Framework for Decision Making. Harvard Business Review, 85(11), 68–76.
  • Torres, T. (2021). Continuous Discovery Habits: Discover Products That Create Customer Value and Business Value. Product Talk LLC.

🔒 Restricted Poll

Please log in to your Scrum Manager account to do this.

Log in
Share:

Comments (0)