Every product development organization has some version of a governance process. Stage gates, go/kill reviews, portfolio committees — the names vary, but the structure is similar: at defined intervals, leadership reviews the status of development programs and decides whether to continue investing in them.
In theory, this is how organizations ensure that resources are allocated to the right programs and withdrawn from the wrong ones. In practice, governance processes in product development fail with remarkable consistency — not because the process is poorly designed, but because the human dynamics around it are almost never accounted for in the design.
The Go/Kill Decision Is the Hardest Decision in Product Development
The most consequential governance moment in any development program is the decision to stop it. To acknowledge that the investment made to date is not recoverable, that the path forward will not produce the outcome the organization originally committed to, and that the resources currently allocated to this program should be redeployed elsewhere.
This decision is almost universally made later than it should be. The reasons are structural and human simultaneously.
On the structural side: most governance processes evaluate programs against the plan they were approved on. But a program can meet every milestone and still be the wrong program — because the market has shifted, the competitive landscape has changed, or the technical assumptions that underpinned the business case have proven incorrect. Milestone completion is not the same as strategic fitness.
On the human side: every program that reaches a stage-gate review has people attached to it — people who championed it, who have spent months or years working on it, whose professional identity is in some measure tied to its success. The organizational cost of killing a program is not just financial — it is social and political. And in most organizations, those costs are borne by the people who make the kill decision, which creates a powerful structural incentive not to make it.
What Governance Processes Consistently Get Wrong
The failure modes in NPD governance cluster around four patterns.
The first is reviewing status instead of fitness. Status tells you where a program is against its plan. Fitness tells you whether the program still deserves to be in the plan. Most governance reviews are structured to evaluate status. The harder question — given everything we now know, should we still be investing in this program? — is rarely asked with the directness it requires.
The second is conflating the program with its champion. Programs championed by senior or influential people receive more benefit of the doubt than programs with less senior champions, independent of their actual merit. This is a natural human dynamic, and it is corrosive to good portfolio management.
The third is insufficient information at the point of review. Governance decisions made without current, accurate, and honestly presented information are decisions made in the dark. The problem is often not that data does not exist — it is that the data presented at reviews is optimistic, incomplete, or structured to support continuation rather than enable genuine evaluation.
The fourth is unclear decision authority. When it is not clear who has the authority to make a go/kill decision, decisions get deferred — to the next review cycle, then the next — while the program continues consuming resources.
What Good Governance Actually Looks Like
Effective NPD governance is not more governance. It is clearer governance — structured to produce decisions rather than document status.
The first requirement is defined criteria for continuation. A program should not proceed from one phase to the next unless it has met specific conditions — not just milestone completion, but evidence that the strategic rationale remains valid, the financial case still holds, and the technical risks have been resolved to a level consistent with the next phase of investment.
The second requirement is honest information. Governance bodies that consistently receive optimistic status reports will consistently make poor decisions. The organizational culture that produces honest upward communication is hard to build and easy to destroy.
The third requirement is independence in the decision. The people who review and approve programs should not be the same people who championed them. Internal technical and readiness reviews conducted by people with no stake in the outcome provide a layer of independent evaluation that catches problems the program team may be too close to see.
The fourth requirement is a genuine kill culture. Organizations that have never killed a program in a stage-gate review do not have a governance process. They have a status reporting process with an approval stamp at the end. The willingness to make a stop decision — implemented decisively and treated as sound portfolio management rather than failure — is what separates governance that works from governance that performs.
The Broader Implication
NPD governance is not a process problem. It is a leadership problem. Organizations that govern their development portfolios well do not have fewer failed programs than others. They have failures that are less expensive — because they are identified earlier, stopped sooner, and the resources they were consuming are redeployed to programs that deserve them.
MKA Insights works with life sciences and bioprocessing organizations on NPD governance design and portfolio management. Contact us.