An organisation that has never declined to proceed at a gate does not have gates; it has scheduled meetings that use the vocabulary of gates.
The decision looks routine. A program reaches the end of its planning phase, the pack is circulated, the committee convenes, the sponsor presents, questions are asked, and approval to proceed is granted.
Nothing about the meeting suggests it was consequential. The material was thorough. The challenge was genuine. The outcome was never seriously in doubt — and that last observation is the one worth sitting with, because it is true of nearly every gate review in nearly every large organisation, and it means the gate is not doing the thing gates exist to do.
A stage gate is not a checkpoint. It is the moment at which an enterprise decides whether to continue committing capital and capacity to an initiative, in full knowledge of what it now knows and did not know when it started. Its defining feature is not the evidence it requires. It is the availability of the answer no.
When that answer is not genuinely available, the entire apparatus — the phases, the artefacts, the committee, the approval — continues to operate at considerable cost while providing none of the control it appears to provide.
The Strategic Context
Almost every delivery framework arranges work into sequential phases with boundaries between them. A common structure uses four: an initiation phase, a planning phase, an implementation phase and a closure phase. Each has its own artefacts and its own decision point.
The structure is sound, and the reasoning behind it is genuinely commercial. Capital is committed progressively rather than all at once, so that the organisation can stop before spending the majority of the money. The boundaries exist to create decision points where uncertainty has reduced enough to make a better judgement than was possible at the start.
That is the theory, and it is a good one. What determines whether it operates is not the framework but a set of organisational conditions almost never specified alongside it: whether the decision-makers can decline, whether declining has a route, and whether anyone's position is made worse by exercising it.
What Leaders Commonly Misread
That evidence quality is the constraint. Gate packs in most enterprises are thorough. The constraint is almost never the quality of information; it is that stopping has no institutional pathway. There is no template for a decision to discontinue, no defined process for redeploying the team, no established way for a sponsor to have made the recommendation without it reading as personal failure.
That a gate is a project control. It is a portfolio control that happens to be exercised on a project. The question at a gate is not whether this initiative is going well. It is whether continuing to fund it remains the best available use of the capital and capacity it consumes — which cannot be answered by looking only at the initiative.
That sunk cost is the main obstacle. Sunk cost bias is real but overstated as an explanation. The more powerful obstacle is that in most organisations, a stopped initiative is recorded as a failure and an initiative that runs to completion and delivers little is recorded as a delivery. The accounting is symmetric; the reputational treatment is not, and people respond to the reputational treatment.
That more gates means tighter control. Gates without genuine optionality add cost and latency without adding control. Three gates at which stopping is real provide more governance than eight at which it is not.
Reframing the Issue
The reframe: a gate is a capital re-allocation decision, and it should be conducted like one.
At each gate the enterprise is deciding, with better information than it had before, whether to commit the next tranche of funding and capacity to this initiative rather than to something else. That framing changes the composition of the pack, the identity of the decision-maker, and the standard against which the initiative is judged.
It changes the pack because the relevant comparison is not against the original business case but against current alternatives. An initiative that is performing exactly as forecast may still be the wrong place for the next tranche if better opportunities have emerged — and the original business case cannot reveal that.
It changes the decision-maker, because a committee constituted around a single initiative cannot make a portfolio comparison. It has no visibility of the alternatives and no authority over them.
And it changes the standard. The question stops being has this program met its commitments? and becomes would we fund this now, at this cost, knowing what we know? Those questions have different answers surprisingly often, and the gap between them is where portfolio value is lost.
A hypothetical software platform program makes this concrete. Two years in, delivery is broadly on plan. But the market has moved, a capability that was going to be built is now available commercially at a fraction of the cost, and the internal capability the program consumes is the same capability three newer initiatives require. Against its own business case, the program passes. Against the portfolio question, it should probably be substantially rescoped. Only the second question surfaces that, and only a forum with portfolio visibility can ask it.
An Instructive Piece of Evidence
One small piece of documentary evidence shows how much the framing of a phase can shift what people do inside it — and how easily the framing drifts back towards completeness.
Two versions of the same teaching material for the same unit, separated by several years, describe the same four-phase lifecycle using different labels. [SOURCE DETAILS REQUIRED: the two documents are versions of one university unit's introductory material; the direction of revision is established by embedded file properties dating the earlier version to 2006 and the later to 2013, not by the text of the documents themselves.]
The earlier version names the phases Project Initiation, Project Planning, Project Implementation and Project Closure. The later version renames them Starting the Project, Organising and Preparing, Carrying out the Project Work and Closing the Project.
The four-phase structure beneath the labels is unchanged, and so is the breakdown of topics within each phase. What moved was the grammar — from abstract nouns to activity phrases. Other revisions were made to the same material at the same time, covering reading requirements and assessment tasks; the phase renaming is the one that bears on governance.
The revision is a small editorial choice, but it points at something structural. Abstract nouns name states the work is in. Activity phrases name what people are doing. And the framing an organisation uses for its phase boundaries influences what a boundary review becomes.
Where a phase is a state, the boundary is a transition between states — and the natural question at the boundary is have we finished this state? That question has a yes or no answer about completeness, and completeness is assessed against a checklist.
Where a phase is an activity, the boundary is the end of one kind of work and the start of another — and the natural question becomes should we now begin doing the next thing? That question is not about completeness. It is about whether the next commitment is warranted, which is a materially different and considerably more useful question.
Neither labelling forces the right behaviour. But organisations that describe their gates in the language of phase completion should not be surprised when their gate reviews assess completion, and when the answer no becomes structurally unavailable — because completeness is nearly always achievable if enough effort is applied, and a committee presented with a complete pack has been given no grounds on which to stop.
Decision Framework
Six tests for whether a gate is genuine. A gate failing three or more is a status review.
| Test | Question | Failure signal |
|---|---|---|
| Optionality | Has this forum ever declined to proceed? | No precedent in three years |
| Comparison | Is the initiative assessed against current alternatives? | Assessed only against its own baseline |
| Standing | Does the decision-maker have authority over the alternatives? | Committee sees one initiative only |
| Pathway | Is there a defined process for stopping? | No template, no precedent, no plan |
| Consequence | What happens to a sponsor who recommends stopping? | Career cost — the decisive failure |
| Timing | Does the gate occur before the next tranche is spent? | Approval follows commitment |
The Consequence test is the one that determines the others. An organisation can build every mechanism on this list and still have no functioning gates if recommending discontinuation is professionally damaging. Conversely, an organisation where stopping well is visibly rewarded will make good gate decisions with quite modest machinery.
From Strategy to Execution
Immediate. Count the gate decisions taken in the last three years and identify how many resulted in anything other than approval to proceed. If the number is zero or near it, the finding is not that the portfolio is uniformly healthy.
Medium term. Move gate decisions for material initiatives into a forum with portfolio visibility, and require that gate papers include the alternative uses of the capital and capability being requested. Without the alternatives, the comparison cannot be made, whoever is in the room. [Related article: Measuring an Outcome You Cannot Predict]
Long term. Build the pathway and the precedent for stopping. Define how a team is redeployed, how partial value is captured, how the decision is recorded, and — most importantly — establish through visible example that a sponsor who recommends discontinuation early has done the organisation a service. Until that is demonstrated rather than stated, every other reform will be absorbed.
Signals to Monitor
- Approval rates near one hundred per cent. The single most reliable indicator that gates are ceremonial.
- Gate papers that discuss only the initiative. No alternatives means no comparison means no decision.
- Gates held after the money is committed. Common where procurement or mobilisation runs ahead of approval.
- Rising pack sizes without rising decline rates. The organisation is increasing the cost of the gate without increasing its function.
- Initiatives that were rescoped quietly rather than reconsidered openly. Often a sign the organisation knows the answer but has no route to give it.
Questions for the Leadership Team
- When did we last decline to proceed at a gate — and what happened to the person who recommended it?
- Do our gate papers show what else the requested capital and capacity could fund?
- Would we fund our three largest in-flight initiatives today, at their remaining cost, knowing what we now know?
- Is there a defined process in this organisation for stopping something well?
- Are our gates scheduled before or after the commitments they are supposed to authorise?
Closing Perspective
The purpose of dividing work into phases was never administrative tidiness. It was to give an enterprise repeated, genuine opportunities to change its mind while changing its mind was still affordable.
An organisation that has built the phases, the artefacts and the committees but not the capacity to say no has bought the entire apparatus of staged investment and retained none of its value. The gate is not the pack, and it is not the meeting. It is the availability of a different answer.
Related article: Measuring an Outcome You Cannot Predict
Related article: Drivers, Supporters and Observers: Who Is Allowed to Change What Your Program Is For
Related article: Whose Knowledge Does Your Governance System Actually Hear?
About the author
Kevin Jogin is Founder & Principal Advisor at EraNorth. Meet the Founder.