AIPOM Investment Stage Gates
What Is It
Create a small set of recurring investment decisions for AI bets, with evidence proportionate to each stage and explicit continue, pivot, scale, pause, stop, and exception paths.
Why Use It
Without gates, pilots persist because no one owns the stopping decision. With bureaucratic gates, teams optimize documents instead of learning. A useful gate names the decision, minimum evidence, accountable authority, and consequence of not meeting the bar.
When to Use It
Use after triage and bet chartering, before multiple initiatives compete for continuing or scaling capital. Tailor rigor to consequence and reversibility rather than imposing one checklist on every experiment.
What It Produces
- Stages and decisions from exploration through retirement
- Minimum evidence and critical controls at each gate
- Decision authority, timing, exception, and appeal rules
- Standard decision record and portfolio learning loop
Who Should Participate
Include portfolio and funding authorities, Product Operations, product and technology leaders, finance, and governance partners. Initiative owners supply evidence but should not approve their own exceptions.
Evidence to Bring
Bring real initiative histories, decisions, costs, evaluations, incidents, dependencies, adoption evidence, and stopped or stalled pilots. Design around actual uncertainty, not hypothetical perfection.
How to Do It
- Define the stages and the investment decision at each boundary.
- Separate exploration evidence from validation and scale evidence.
- Specify required outcome, behavior, feasibility, economic, adoption, and responsibility evidence.
- Identify critical gaps that block progression regardless of other strengths.
- Assign decision authority, contributors, reviewers, and escalation.
- Define continue, pivot, scale, pause, stop, and retire criteria.
- Create time-bounded exceptions with an owner, rationale, compensating control, and expiry.
- Set review cadence and capture what portfolio criteria should change from experience.
- Test the model against past initiatives and revise ambiguous gates.
Key Concepts
- Gates fund the next uncertainty, not the original story.
- Scale requires representative evidence and operating controls.
- Critical gaps are not averaged away.
- An expired exception becomes a decision, not permanent ambiguity.
Organizational Applications
Use for innovation funds, quarterly planning, platform investments, vendor-backed pilots, and retirement of low-learning initiatives.
Common Pitfalls
- Requiring the same evidence at every stage
- Approving documents instead of examining use
- Hiding stop decisions behind endless pilots
- Treating sunk cost as evidence
- Allowing initiative owners to approve exceptions
- Scaling before production, workflow, and responsibility evidence exists
Combine With
Use aipom-portfolio-posture-advisor to diagnose portfolio balance, aipom-economic-case-builder for material scale economics, and aipom-initiative-readiness-review before consequential launch or expansion.
Assets and Templates
- Stage-gate design template
- Synthetic worked example
- Weak example
Sources
This skill is an original AIPOM synthesis of evidence-based staged investment and portfolio decision practice.
1---2name: aipom-investment-stage-gates3description: Define evidence-based entry, continuation, pivot, scale, pause, and stop decisions for AI investments without turning gates into document approval theater.4---56# AIPOM Investment Stage Gates78## What Is It910Create a small set of recurring investment decisions for AI bets, with evidence proportionate to each stage and explicit continue, pivot, scale, pause, stop, and exception paths.1112## Why Use It1314Without gates, pilots persist because no one owns the stopping decision. With bureaucratic gates, teams optimize documents instead of learning. A useful gate names the decision, minimum evidence, accountable authority, and consequence of not meeting the bar.1516## When to Use It1718Use after triage and bet chartering, before multiple initiatives compete for continuing or scaling capital. Tailor rigor to consequence and reversibility rather than imposing one checklist on every experiment.1920## What It Produces2122- Stages and decisions from exploration through retirement23- Minimum evidence and critical controls at each gate24- Decision authority, timing, exception, and appeal rules25- Standard decision record and portfolio learning loop2627## Who Should Participate2829Include portfolio and funding authorities, Product Operations, product and technology leaders, finance, and governance partners. Initiative owners supply evidence but should not approve their own exceptions.3031## Evidence to Bring3233Bring real initiative histories, decisions, costs, evaluations, incidents, dependencies, adoption evidence, and stopped or stalled pilots. Design around actual uncertainty, not hypothetical perfection.3435## How to Do It36371. Define the stages and the investment decision at each boundary.382. Separate exploration evidence from validation and scale evidence.393. Specify required outcome, behavior, feasibility, economic, adoption, and responsibility evidence.404. Identify critical gaps that block progression regardless of other strengths.415. Assign decision authority, contributors, reviewers, and escalation.426. Define continue, pivot, scale, pause, stop, and retire criteria.437. Create time-bounded exceptions with an owner, rationale, compensating control, and expiry.448. Set review cadence and capture what portfolio criteria should change from experience.459. Test the model against past initiatives and revise ambiguous gates.4647## Key Concepts4849- Gates fund the next uncertainty, not the original story.50- Scale requires representative evidence and operating controls.51- Critical gaps are not averaged away.52- An expired exception becomes a decision, not permanent ambiguity.5354## Organizational Applications5556Use for innovation funds, quarterly planning, platform investments, vendor-backed pilots, and retirement of low-learning initiatives.5758## Common Pitfalls5960- Requiring the same evidence at every stage61- Approving documents instead of examining use62- Hiding stop decisions behind endless pilots63- Treating sunk cost as evidence64- Allowing initiative owners to approve exceptions65- Scaling before production, workflow, and responsibility evidence exists6667## Combine With6869Use `aipom-portfolio-posture-advisor` to diagnose portfolio balance, `aipom-economic-case-builder` for material scale economics, and `aipom-initiative-readiness-review` before consequential launch or expansion.7071## Assets and Templates7273- [Stage-gate design template](template.md)74- [Synthetic worked example](examples/worked-example.md)75- [Weak example](examples/weak-example.md)7677## Sources7879This skill is an original AIPOM synthesis of evidence-based staged investment and portfolio decision practice.
Dean Peters