Vizually
ArticleProject Lifecycle3 min read

5 Go/No-Go Mistakes Executives Make on Enterprise Software Programs

Five specific go/no-go mistakes enterprise software executives make when estimation patterns are screaming the wrong answer. A concise corrective for portfolio leaders.

Vizually Team·
Initiation & Chartering

Five expensive go/no-go reflexes — and what to do instead

Most enterprise software no-go decisions arrive six months too late and cost ten times what an earlier conversation would have cost.
Vizually editorial

Enterprise software programs accumulate enough sunk cost by gate three that executives systematically tilt toward go decisions. The estimation pattern — forecasts wrong in the same direction — is most expensive when an executive layer doesn't see it clearly enough to halt a program that's drifted past viability. The five mistakes below are the most common reflexes that produce that outcome.

Each is paired with a corrective short enough to use at the next gate.

  1. Mistake 1
    Treating the gate as a yes-or-no when conditional-go is the right answer
    Binary gates encourage teams to over-promise to clear them. Conditional-go with named conditions and a re-review date produces honesty. Corrective: replace 'go/no-go' with 'go / conditional-go / no-go' as the standard option set.
  2. Mistake 2
    Discounting the team's own confidence intervals
    When the team gives a range, executives anchor on the optimistic end. Corrective: gate decisions on the worst-case end of the team's range, not the best-case. If the program isn't viable on the worst case, the gate is a no-go.
  3. Mistake 3
    Greenlighting on momentum, not on data
    Once a program has cleared two gates, the third gate is treated as a formality. Corrective: every gate has the same evidence requirement as gate one. If the team can't produce that evidence at gate three, that's the answer.
  4. Mistake 4
    Ignoring the parallel program competing for the same engineering capacity
    Software gates are made program by program; resource conflicts are made portfolio by portfolio. Corrective: every gate decision shows the engineering capacity assumption in writing, sourced from the engineering org, not the program team.
  5. Mistake 5
    Not pre-defining what would change a go to a no-go
    Without a tripwire, every program continues until it's obviously failed — at which point the cost of stopping has dwarfed the cost of an earlier pause. Corrective: every go decision documents one or two specific tripwires that would convert it to no-go. The tripwires are reviewed at the next gate.

Three-decision discipline at every gate

0 / 5
  • Is the option set 'go / conditional-go / no-go' rather than binary?
  • Are we evaluating the worst case of the team's range, not the best case?
  • Are we reviewing fresh evidence rather than relying on prior gate decisions?
  • Is the engineering capacity assumption sourced from the engineering org?
  • Have we documented at least one tripwire that would convert today's decision?
MistakeWhat it producesThe corrective
Treating go/no-go as a milestone, not a decisionCalendar drives the gateRe-pose the question: "what would make us not ship?"
Silent stakeholder absent from gateLate vetoPre-gate conversation with veto-holders, in writing
No reversibility planShip-or-die mindsetPre-document what a controlled rollback looks like
Estimation pad burned by Q3Final approvals are unsafeReset the pad mid-project, not at the gate
Conflating "ready" with "complete"Half-built features shipDefine ready as the smallest releasable scope

These five corrections are not difficult, and they don't require a heavier gate process. They require an executive willing to apply discipline that the program team cannot apply for them. The corrective intervention here is short — five mental checks before voting at the next gate.

For the same pattern viewed at greenlight rather than mid-gate, see the preventive greenlight question; for the annual retrospective that builds the data behind these checks, see the retrospective playbook.

More in
CategoryProject Lifecycle

Related reading

Articlethe preventive greenlight question for creative workArticlethe annual retrospective playbookArticlethe hardware checklist version