The Initiative Trap
Why successful change initiatives don’t always translate into better performance.
A transformation can succeed initiative by initiative and still disappoint as a transformation. The technology launches. The process improves. Adoption rises. AI use cases move into production. By the measures attached to each effort, progress is real. Yet the business outcome that justified all of it—greater productivity, faster growth, lower cost, shorter time to market—barely moves. Sometimes the problem is execution. But sometimes the initiatives worked exactly as intended. They simply didn’t add up to the outcome. That is the initiative trap.
Local success can disappear in the system
Consider a marketing organization trying to reduce campaign time to market. One team introduces generative AI and cuts content production from six days to two. Another implements a new work-management platform. A third improves intake. Each initiative succeeds. But legal review still takes eight days. Priorities change after work begins. Teams receive incomplete briefs. Multiple approval layers remain. Content production becomes four days faster. The campaign does not.The improvement was real. It was simply absorbed by constraints elsewhere in the system.
That is because outcomes such as speed, productivity, growth, and customer experience are rarely produced by one technology, process, or team. They emerge from the way decisions, workflows, technology, roles, capabilities, and behaviors work together.
Management research has described this through the idea of organizational complementarity: the value of one organizational change can depend on the choices surrounding it. Research on workplace practices and technology adoption has similarly found that combinations of complementary changes can produce different results from isolated improvements.
The management implication is simple:Improving part of a system does not guarantee that the system improves.
Initiatives solve one problem—and can create another
Organizations need initiatives. They create ownership, scope, budgets, milestones, and accountability. Without them, complex change quickly becomes unmanageable. But business outcomes do not follow the same boundaries. A technology team owns the platform. An AI team owns use cases. A process team owns workflow. A business leader owns performance. Meanwhile, the work producing that performance moves across all of them.
We organize accountability vertically. Performance is produced horizontally.
The trap begins when the structure used to deliver change becomes the structure used to design and manage the transformation itself. Then one business outcome becomes five initiatives, each with its own plan and definition of success. Everyone can succeed at their piece. No single piece can produce the result.
Better coordination is not enough
The usual answer is to coordinate the initiatives more closely. That helps. But six independently designed initiatives do not become one coherent transformation simply because their leaders meet more often.
If each initiative begins with its own solution, scope, and measures, integration is already arriving late. A better approach reverses the sequence. Start with the outcome.Then follow the work across the organization.
Where does it slow down? Where are decisions revisited? Where do handoffs break? Where do priorities collide? Where has an improvement in one place merely exposed the next constraint? Only then decide what needs to change. When initiatives are the starting point, integration happens after solutions have been defined. When the outcome is the starting point, integration becomes part of the design.
See the whole. Change the critical few.
Taking a systems view does not mean creating a massive transformation. Quite the opposite.The purpose of seeing the whole system is to determine what does not need to change.A useful principle is:
See the whole system. Change the critical few.
The critical few are the smallest set of interdependent changes sufficient to materially move the outcome. For one organization, that might mean changing a workflow, two decision rights, an approval structure, and the technology supporting them. For another, the technology may already be adequate. The real constraints may be prioritization and governance.
A broader diagnosis can lead to a narrower transformation. That becomes even more important as AI allows organizations to improve individual activities faster. Faster content creation, analysis, automation, or decision support can create real value—but those gains can disappear if the surrounding work does not change with them.
As the capacity for local change increases, integration becomes a more important management capability.
Manage transformation through outcomes
Organizations will always need initiatives. They remain one of the best ways to organize and deliver change. But initiatives are means.Performance is the point.
Before launching the next one, leaders should ask three questions:
What performance must move?
What combination of constraints is holding it in place?
What is the smallest set of things that must change together?
Then use initiatives to deliver those changes. Deliver through initiatives. Manage transformation through outcomes.
That is how organizations avoid succeeding at change while falling short on performance.And how they escape the initiative trap.
Research foundations
Paul Milgrom and John Roberts, “Complementarities and Fit: Strategy, Structure, and Organizational Change in Manufacturing,” Journal of Accounting and Economics, 1995.
Casey Ichniowski, Kathryn Shaw, and Giovanna Prennushi, “The Effects of Human Resource Management Practices on Productivity,” American Economic Review, 1997.
Timothy F. Bresnahan, Erik Brynjolfsson, and Lorin M. Hitt, “Information Technology, Workplace Organization, and the Demand for Skilled Labor,” Quarterly Journal of Economics, 2002.