A practitioner-ready product lifecycle framework
that treats time, kills, and named authority as load-bearing structure.
The Complete Operating Manual
Most PLM frameworks assume a committee at every phase keeps bad products out. It doesn't. What it keeps out is speed, ownership, and the honest reckoning a product team needs to make a kill decision without political cover.
Kill criteria, clocks, and named owners do the real work. FORGE keeps two formal gates. Everything between them is the product owner's problem.
Most products die of oxygen deprivation, not poor decision-making. The clock, the kill, the named owner, and the Furnace are the corrections. Everything else is commentary. Every other framework treats the organization as the asset and the process as its protector. FORGE starts from the opposite position: the product is the asset, and the organization is the thing most likely to suffocate it.
This framework was built inside a top-three national payer's product organization after 50+ interviews with product owners, clinical leaders, compliance teams, and executive sponsors who described the same failure modes. The phases are not theoretical. The kill criteria are not aspirational. The people mechanisms are not borrowed from a textbook. Every element exists because something specific broke without it.
Not every product needs the full pipeline. Classification determines phase count, deliverable load, gate requirements, and time box. Get it right at the start or the framework punishes you later.
The complete pipeline. New product, new market position, new infrastructure. No shortcuts.
New capability on an existing product. Inherits parent product infrastructure, lighter deliverable load, compressed bookends.
Iteration, fix, or enhancement. Product owner has full authority. No gates. If scope grows beyond the time box, reclassify as Function and trigger Gate 1.
If an Update exceeds its time box, it must be reclassified as a Function. This triggers Gate 1 review. No exceptions, no time-box extensions.
Acquired product absorbed into portfolio. The acquisition decision was Gate 1. FORGE picks up at Build.
Classify at intake. The matrix determines everything downstream.
| Dimension | Full | Function | Update | Integration |
|---|---|---|---|---|
| Creates new product? | Yes | No | No | Acquired |
| Creates new capability? | Yes | Yes | No | Inherited |
| Council involvement? | Gate 1 + Gate 2 | Gate 1 + Gate 2 | None | Gate 2 only |
| Max time box | 40 weeks | 20 weeks | 7 weeks | 24 weeks |
| Phases | Signal → Thesis → Build → Prove → Launch → Sustain | Thesis → Build → Prove → Launch → Sustain | Build → Prove → Release | Build → Prove → Launch → Sustain |
| Deliverable load | Full | Inherited + delta | Minimal | Integration-specific |
| Reclassification trigger | N/A | N/A | Exceeds time box → Function | Cost >30% → Kill review |
The council sees the product twice. Every meeting that is not a gate is the product owner's responsibility to schedule or skip. Gates are binary: the product proceeds or it doesn't.
Should we commit build resources to this product?
5–7 people: executive sponsor, clinical lead, finance, operations, technology, 1–2 rotating peers
20 min presentation + 20 min questions + 5 min decision = 45 min max
Proceed: build resources committed. Kill: permanent. Send to Furnace: the idea has potential but the thesis is not ready for the Pipeline.
Should we commit market-facing resources to this product?
Gate 1 council + sales/channel lead + marketing lead + legal sign-off
30 min presentation + 20 min questions + 10 min decision = 60 min max
Documented buyer willingness to pay. Not intent. Not interest. Willingness to pay, in writing.
Proceed: market resources committed. Kill: permanent. Pivot: product scope redefined, re-enters at Thesis. All sunk cost absorbed by originating BU.
Same deliverables as gates, different reviewer. The product owner signs off. No meeting required. No scheduling delay. The checkpoint creates a decision record. That is all it does.
Third attempt at the same phase triggers mandatory council escalation. The PO gets two tries. After that, the decision leaves their hands.
These are neither status meetings nor optional, and they carry the same standard as gates. Same standard, different signer. The difference is who decides.
Where ideas live before anyone commits build resources. Protected budget, separate owner, no obligation to ship.
Separate from the Pipeline: different owner, different budget, different expectations. If the Pipeline is the engine, the Furnace is the wind tunnel. Nothing in the Furnace is obligated to graduate. Nothing in the Pipeline can raid Furnace resources.
Open-ended investigation. No deliverable obligation. Budget caps it, not phases.
Structured experiment with hypothesis and success criteria. Closest analog to Pipeline's Build/Prove, but without the pipeline clock.
Pulling a component out of an existing product for standalone development. The parent product continues unaffected.
Exploring a market or capability adjacent to current portfolio. Not a product yet. Might become a Signal.
Long-range market and technology monitoring. No build activity. Output is a brief, not a backlog.
Structured retrieval of learnings from killed products. Within 30 days of a Pipeline kill, two senior product people can petition the Furnace Lead. One 90-day cycle, non-renewable. The question: did the kill criteria miss something? Recovery succeeds about 20% of the time. The other 80% confirms the kill was right, which matters more than it sounds.
of org capacity
Protected allocation
Capacity review
Three exit paths: graduate to Pipeline at Signal or Thesis phase, kill, or persist in current mode. There is no pressure to graduate. The Furnace earns its allocation by answering questions the Pipeline can't ask yet.
The Pipeline is the inhale: structured phases, binding kill criteria, hard clocks. The Furnace is the exhale: protected exploration, political extraction, long-range scanning. An organization that only inhales suffocates on its own rigor.
The Furnace's self-correction impulse comes from the same place as Mechanism 15. Both are rooted in The Organizational Pendulum — the observation that every system drifts toward either rigidity or chaos, and the ones that survive build detection into their own structure.
The Pipeline has six phases, each with named deliverables, kill criteria, and time boxes. The Furnace needs the same discipline. Without per-mode specifications, the Furnace becomes a parking lot for ideas nobody wants to kill.
Pure signal detection. No thesis required. Looking for problems worth naming.
REQUIRED DELIVERABLESNo named problem with an identifiable buyer after the time box expires. Also kills if the problem is already being solved by a Pipeline product.
A named problem with at least three independent signal sources and a buyer type willing to discuss it. Enough to enter Incubation.
Early-stage concept development. Hypothesis exists but is untested.
REQUIRED DELIVERABLESHypothesis cannot be validated within the time box. Buyer conversations produce indifference. Also kills if a structural barrier makes the concept unworkable.
A validated hypothesis with buyer evidence and a recommendation to extract into Pipeline.
Pulling a proven Furnace concept into Pipeline for productization.
REQUIRED DELIVERABLESNo Pipeline capacity within 30 days. Also kills if the CPO determines the evidence does not meet Pipeline Thesis standards.
Product enters Pipeline at Thesis with a named PO, locked scope, and Furnace evidence as foundation. 30-day Transition Quarantine activates.
Extending existing products into neighboring use cases.
REQUIRED DELIVERABLESThe adjacent market requires more than 30% modification to the existing product. At that point it is a new product and belongs in the Pipeline as a de novo entry.
Extend the existing product (Function change) or fork a new product (enters Pipeline at Thesis via Extraction).
Long-range market, regulatory, and technology monitoring.
REQUIRED DELIVERABLESTwo consecutive quarters with no actionable signals and less than 30% prediction accuracy. Pause for one quarter and reassess scope.
Horizon Scanning does not exit. It produces inputs. When a signal crosses the action threshold, it triggers an Exploration cycle.
Structured retrieval from killed products. Single cycle, no extensions.
REQUIRED DELIVERABLESThe evidence test confirms the original kill was correct. Also kills if the 90-day clock expires without a clear recommendation.
Product re-enters Pipeline at Thesis with new evidence and a new product owner, or the kill is confirmed and the archive is updated.
The Furnace Lead owns the Furnace the way a Product Owner owns a Pipeline product. One person. Named, not titled. A senior product leader who reports directly to the CPO and operates outside the Pipeline reporting structure.
Furnace Lead reports to CPO. Not to a Pipeline Product Owner, not to a BU leader, not to an innovation committee. Direct line to the person who owns the portfolio.
Furnace Lead vs. Pipeline Product Owner conflicts: CPO resolves within 5 business days. If the clock expires, the Furnace Lead's position holds. The Furnace is structurally protected because it is structurally weaker.
15–20% of total product budget, ring-fenced. Set annually, reviewed quarterly, cannot be raided for Pipeline overruns. The moment the Furnace budget becomes a Pipeline reserve fund, the Furnace is dead.
The Pipeline has council gates. The Furnace needs governance too, but lighter. Two cadences.
30 minutes. Furnace Lead presents to CPO and Oversight Group. Status of every active cycle: on track, at risk, or recommending kill. Budget burn vs. plan. Any cycles requesting entry or recommending Extraction. If no decisions are needed, the review takes 15 minutes.
Format: Written update 2 days before. Meeting covers only decision items. If none, the meeting is canceled. Attendance: Furnace Lead, CPO, Oversight Group (2–3 people). No observers.
The formal gate for moving a Furnace concept into Pipeline. Triggered by the Furnace Lead when a cycle produces a positive recommendation. Not scheduled. Called when needed. The CPO reviews the extraction brief, scope lock, and transfer plan.
Decision authority: CPO. One person, one decision, one accountability.
FORGE rests on three structural legs. Remove any leg and the framework collapses into process theater. The Furnace sits at the center, the seat the three legs support.
The seat the three legs support. Exploration gets protection. Discipline does not mean rigidity.
Process does not solve people problems. Structure does. Each one exists because a specific thing breaks without it.
Core decision architecture. These six are non-negotiable in any FORGE deployment.
Added in v2.0. Each one comes from watching the original six fail under specific conditions.
These three modify behavior in the original six without adding new process.
Capacity limits are hard. Four products in three slots means all four ship late.
max Full products in Build/Prove simultaneously
max per parent product
10–15% resource cap
Max three Full products in Build or Prove simultaneously. When the pipeline is full, new products queue at Thesis. One person owns the queue.
The queue is ordered by market urgency. External deadline beats competitive pressure, which beats internal demand. No portfolio review committee. A list, owned by a name, updated weekly.
Five decisions belong to the CPO. Not a committee. Not a working group. One person with the authority and the accountability to make these calls on a weekly cadence.
Concrete constraints. Not guidelines. Not recommendations. Rules that trigger specific consequences when violated.
What the CPO monitors. Not vanity metrics. Signals that the system is working or drifting.
| Indicator | What It Measures | Alarm Threshold |
|---|---|---|
| Pipeline Velocity | Time from Signal to Launch, segmented by product type. Velocity increasing without quality degradation is healthy. | Average velocity exceeds time-box sum by 40%+ |
| Furnace Conversion Rate | Percentage of Furnace cycles that produce Pipeline candidates. Healthy range: 15–30%. | Below 10% or above 40% for two consecutive quarters |
| Kill Rate | Percentage of products killed before Launch. Healthy: 40–60%. Below 30% means gates are rubber stamps. | Below 25% for two consecutive quarters |
| Resource Utilization | Pipeline-to-Furnace split against 85/15 target. Also tracks idle capacity. | Furnace allocation drops below 10% of total budget |
| Gate Pass Rate | Percentage passing each gate. Healthy: Gate 1 at 50–70%, Gate 2 at 60–80%. Gate 1 should be the harder filter. | Gate 1 pass rate exceeds 85% for three consecutive reviews |
| Queue Age | How long products sit in queue before entering Build. Growing queue with stable capacity signals Signal-phase filtering needs tightening. | Any product in queue longer than 6 months without a gate decision |
How products move through the queue. Explicit rules that remove ambiguity about what happens next.
Modify anything else in FORGE to fit your organization. Do not modify these.
Required deliverables gate the next phase. Optional deliverables are there when the product needs deeper scrutiny. The product type determines which columns apply.
| Phase | Time Box | Required Deliverables | Optional | Full | Function | Update | Integration |
|---|---|---|---|---|---|---|---|
| Signal | 2 wks | Signal memo (1 page: market signal, size estimate, strategic fit) Competitive scan (who else is here?) |
Customer interview notes | Yes | — | — | — |
| Thesis | 4–6 wks | Product thesis (problem, buyer, differentiation, unit economics) Kill criteria (falsifiable, written) Buyer evidence (3+ conversations documented) |
Financial model draft Technical feasibility memo |
Yes | Compressed | — | — |
| Gate 1 | 45 min | Gate 1 deck (per format rules) Kill criteria review Resource request |
Red team report | Yes | Yes | — | — |
| Build | 8–12 wks | Working prototype or MVP Technical architecture doc Updated kill criteria Checkpoint memo |
Integration spec Security review |
Yes | Yes | Compressed | Yes |
| Prove | 8–12 wks | Pilot results with defined success metrics Buyer willingness to pay (documented) Operational readiness assessment Checkpoint memo |
Scale plan Regulatory clearance |
Yes | Yes | Compressed | Yes |
| Gate 2 | 60 min | Gate 2 deck (per format rules) Pilot evidence package Go-to-market plan Financial model (final) Legal sign-off |
Pre-mortem output | Yes | Yes | — | Yes |
| Launch | 4–6 wks | Launch plan (channels, messaging, timeline) Sales enablement package Support/ops handoff Success metrics defined |
PR/comms plan Training materials |
Yes | Compressed | — | Yes |
| Sustain | Ongoing | Quarterly health scorecard Annual zombie detection review Scope fingerprint (annual) |
Sunset plan (if triggered) | Yes | Absorbed | — | Yes |
Editable templates for every deliverable, gate review, and governance document in the framework. Each template enforces the deliverable standard for its phase. Interactive versions are available on the FORGE site.
| Template | Phase / Context | Key Fields | Purpose |
|---|---|---|---|
| Signal Brief | Signal | Signal source and credibility, buyer hypothesis, competitive scan, go/kill recommendation | Forces the first honest answer: is this a real signal or internal noise? |
| Thesis Document | Thesis | Hypothesis, market need, target buyer, revenue model, competitive positioning, falsification criteria | The structural case for why this product should exist. Written before anyone writes code. |
| Kill Criterion Writer | All Phases | Metric, threshold, measurement date, named decision-maker, falsifiability quality check | Prevents vague kill criteria that cannot be disproven by available data. |
| Unit Economics Model | Thesis / Build | Revenue per unit, cost per unit (itemized), gross margin, break-even volume, sensitivity analysis | Structural economics, not projections. Names the number where the model breaks. |
| Gate 1 Presentation Package | Gate 1 | Thesis summary, unit economics, competitive positioning, kill criteria for Build, recommendation | Everything the council needs to make a concept approval decision in 45 minutes. |
| Pilot Results Report | Prove | Projected vs. actual metrics (conversion, revenue, cost, satisfaction), buyer feedback, operational findings, recommendation | What we projected. What happened. Whether the delta is fatal or addressable. |
| Gate 2 Presentation Package | Gate 2 | Pilot-validated economics, GTM plan summary, operational readiness, post-launch kill criterion, recommendation | Final go/kill decision with pilot-validated data, not thesis assumptions. |
| Go-to-Market Plan | Launch | Sales targets (named accounts, owners, deadlines), marketing activities, channel strategy, total budget | Named owners. Named deadlines. Named budgets. No abstract launch plans. |
| Quarterly Performance Dashboard | Sustain | Four-quarter trend for financial performance, customer retention, satisfaction, operational cost per unit | Forces the quarterly question: is this product still earning its portfolio slot? |
| Checkpoint Memo | Build / Prove | Phase summary, kill criteria status, deliverable checklist, advance/restart/escalate decision | Same deliverables as gates, different reviewer. Creates a decision record without a meeting. |
| Build Phase Checkpoint | Build | Build objective, milestone progress, kill criteria status, blockers, resource burn rate, continue/pause/kill | Prevents Build from becoming an open-ended engineering project. |
| Furnace Cycle Plan | Furnace (All Modes) | Hypothesis, time box, budget envelope, FTE allocation, kill criterion, expected output, extraction criteria | Furnace work without a hypothesis is tourism. This template enforces discipline. |
| Post-Kill Documentation | All Phases | Kill criterion triggered, total investment (time + dollars), lessons learned, salvageable assets, disposition | Dead products teach more than live ones. Captures institutional knowledge before it evaporates. |
| Extraction Brief | Furnace → Pipeline | Furnace cycle summary, evidence gathered, proposed Pipeline entry phase, resource requirements, risk assessment, sponsor recommendation | The gate between exploration and commitment. Moving to Pipeline requires evidence, not enthusiasm. |
| Portfolio Queue Manager | Portfolio | Active Pipeline products (phase, health, next gate, FTE, kill criteria), active Furnace cycles, resource summary, dependencies, queue decisions needed | CPO-level operating view. Every active product, its phase, its health, and what it costs. |
| Sustain Review Template | Sustain | Performance vs. targets, margin trajectory, competitive position, customer concentration risk, sunset criteria check, renew/invest/harvest/sunset | The quarterly reckoning. Is this product still earning its place in the portfolio? |
| Leadership Handoff Memo | All Phases | Current state, pending decisions, relationship map (champion/neutral/resistant), known risks, institutional knowledge | Everything the next owner needs that is not written down anywhere else. |
Templates are not optional documentation. Each one enforces the deliverable standard for its corresponding phase or gate. A checkpoint without a Checkpoint Memo is a conversation, not a decision record. A Gate 1 without a Presentation Package is a pitch meeting, not a governance event.
The sequence through a gate is deliberate because the order information arrives determines whether evidence or politics drives the outcome.
Strategic alignment comes last because it is the most susceptible to narrative manipulation. The evidence goes first.
Before the Gate 2 decision, each council member independently answers: "It is 12 months from now and this product has failed. What happened?" Responses are collected anonymously and read aloud. The product owner responds to each scenario.
"Would you invest the remaining 20% as new money?" If the council would not fund this product from scratch with the remaining budget, the product should not proceed on momentum alone. This question is asked explicitly at every gate.
Every mechanism in FORGE exists because one of these ten pathologies showed up in an organization I was in. The mechanism is the scar tissue.
Theory is cheap. Here is a single product moving through FORGE from signal to launch, with real numbers and real decisions.
A payer's specialty care team identifies a gap: high-risk maternity members generating avoidable NICU admissions. The clinical signal is real. The cost data supports it. Nobody has built a structured intervention.
3,200 high-risk pregnancies annually. $78K average NICU cost. 22% flagged as avoidable. Three-source signal documented. Go recommendation signed.
Falsification criteria written: product dies if fewer than 3 buyers confirm willingness to pay at $2.50+ PMPM. Five conversations completed. Three confirmed. Unit economics: break-even at 1,800 members.
Council reviews. Thesis holds. Economics credible. Kill criteria falsifiable. Proceed to Build. Clock: 10 weeks.
Working clinical pathway with risk stratification, provider dashboard, member outreach. Operational cost model on actuals. Kill criteria: none triggered.
Pilot with 400 members. 18% reduction in avoidable NICU admissions. Provider adoption 62% vs 80% target. Revised break-even: 2,100 members. Buyer feedback: $2.75 PMPM if adoption improves.
Pilot data reviewed against kill criteria. NICU reduction strong. Adoption gap addressable. GTM plan with named owners. Ops readiness confirmed by the team that will run it, not the team that built it. Proceed to Launch.
The same organization attempted the same product the previous year. It took 14 months and never launched.
Month 1: scope defined. Month 2: undocumented strategy shift. Months 3–5: cycling iterations over what "maternity" means. Month 6: pricing rebuilt for the third time. Month 8: compliance flagged TCPA risk. Month 10: compliance cleared, vendor contract expired. Month 12: new RFP required. Month 14: deprioritized by incoming leadership.
The product was never bad. The structure made speed impossible and nobody had permission to say so.
FORGE against the five frameworks most likely already in the room.
| Dimension | FORGE | Stage-Gate | PACE | Lean PD | SAFe | Pragmatic |
|---|---|---|---|---|---|---|
| Gates | 2 (concept + launch) | 4–6 gates | Phase reviews (variable) | None (continuous) | PI boundaries | None formal |
| Kill criteria | Written, falsifiable, pre-committed | Scorecard-based | Resource-based | Pivot metrics | WSJF-based | Market-based |
| Time discipline | Hard time boxes, no extensions | Flexible timelines | Milestone-driven | Iteration-based | PI cadence (fixed) | Market-driven |
| Parallel exploration | Furnace (protected, non-raidable) | Discovery stage | Platform teams | Build-Measure-Learn | Exploration enabler | Market sensing |
| Portfolio limits | Hard caps (3 Full, 5 Function) | Soft prioritization | Resource allocation | WIP limits | Capacity-based | Market opportunity |
| People mechanisms | 15 named mechanisms | RACI-based | Core team model | Team autonomy | ART structure | Role-based |
| Decision authority | Named person + clock | Gatekeeper(s) | PAC committee | Team/founder | Product mgmt | Product mgr |
| Self-correction | Annual self-diagnostic (Mech. 15) | Post-launch review | Not explicit | Retrospectives | Inspect & Adapt | Win/loss analysis |
| Primary failure mode | Over-rigidity if self-diagnostic skipped | Gate theater, zombie products | Complexity, heavyweight | No structure at scale | Process bureaucracy | Execution gap |
| Phase | Full | Function | Update | Integration |
|---|---|---|---|---|
| Signal | 2 weeks | — | — | — |
| Thesis | 4–6 weeks | 2–3 weeks | — | — |
| Build | 8–12 weeks | 6–8 weeks | 2–3 weeks | 8–12 weeks |
| Prove | 8–12 weeks | 4–6 weeks | 1–2 weeks | 4–6 weeks |
| Launch | 4–6 weeks | 2–3 weeks | Release | 4–6 weeks |
| Sustain | Continuous | Absorbed | — | Continuous |
| Total | 26–40 wks | 14–20 wks | 4–7 wks | 16–24 wks |