CPO Self-Review Partner — PRISM Strategy

Intent

Head of Product/CPO conducts evidence-based self-review of their strategy document to validate: (1) Core Thesis, (2) Strategic Options comparison, (3) Capital Allocation decisions, (4) Pre-mortem/Kill-Switch + Adaptation definition, (5) Mission/Vision alignment and Stakeholder alignment. Outputs: Markdown (human-readable) + JSON (machine-readable) simultaneously generated.

Folder Globs (read-only)

Safe globs reflecting attached structure. External/unconnected materials only referenced when version snapshots exist.

• company-level-context/**/*
• company-level-context/okrs/**/* # includes okr-sparring-partner.mdc
• company-level-context/product-vision-and-strategy/**/* # product-strategy.md, product-strategy-review.mdc
• team-structure/**/* # team structure/role context
• frameworks/continuous-discovery-habits/**/* # interview/opportunity/solution/assumption-testing
• evidence-guided/**/* # calculate-ICE-score.mdc etc.
• guides/product/**/* # create-1-pager.mdc, create-prd.mdc, generate-tasks.mdc, process-task-list.mdc
• initiatives/**/* # individual initiatives, _templates/initiative-template included
• meetings/**/* # meeting/retrospective notes
• .cursor/rules/**/* # related rule references allowed

Inputs (required)

• doc_link (string, required) # strategy document to review
• evidence_hub_link (string, required) # initiatives// or meeting-notes/ etc. root hub
• product_or_area (string, required)
• owner (string, required)
• review_window_start (YYYY-MM-DD, required)
• review_window_end (YYYY-MM-DD, required)
• riskiest_assumptions (string[], minItems=2, required)
• baseline_metric (string|number, optional)
• north_star (string, optional) # NSM
• capacity_window (object, optional) # {fte:number, budget:number}
• runway_months (number, optional)

If missing → HOLD: Return MissingInputs list with 1-Day Evidence Sprint proposal only (no scoring).

Citation Standard (required)

All evidence must use this format:

• [[ev-id|path#line-range|YYYY-MM-DD]] Example) [[ev-12|initiatives/kbo/analysis.md#L42-L77|2025-09-10]]

Missing citations result in 1-step deduction per dimension.

------------------------------

Modes — Self-Review Flow

------------------------------

Mode 0 — Principle Alignment (Mission/Vision/Principles Coherence)

• Mission/Vision Coherence: How does this strategy connect to company Mission/Vision/Principles?
• Vision Narrative (1–3y): One paragraph Narrative of success 1–3 years from now.
• Value Thesis (one sentence): Value mechanism of "Why Us, Why Now".

Mode 1 — Socratic Checklist (Core Questions)

• Thesis (core claim in one sentence)
• Outcome (Leading vs Lagging distinction, customer/business)
• Opportunity (ICP/Segment/Use-case, TAM/SAM/SOM evidence)
• Causality (Why Now? Change factors: market/technology/user behavior/regulation)
• Alternatives (≥3: Do Nothing / Focused Bet / Portfolio Bet)
• Trade-off & Non-goals (what we intentionally won't do)
• Unit Economics (LTV/CAC/Payback/GMs/CMs) key assumptions
• Disconfirming Evidence (counterexamples) and rejection reasons

Mode 2 — Bias Guardrails (Bias Prevention)

Must address each of these 6 in one line each, otherwise Proceed prohibited (=HOLD).

• Confirmation / Availability / Sunk Cost / Survivorship / Overconfidence / Anchoring

Mode 3 — Options & Capital Allocation (Comparison Table)

For each option:

• Assumptions (Top-3) + Confidence (0–1)
• Investment: FTE, CapEx, Opex, Time-to-Impact (weeks)
• Impact: NSM/OKR expected range (Δ)
• Risks (Top-3) + Mitigation
• Non-goals (out of scope)
• Decision Rule (selection threshold/condition; e.g., Payback ≤ 9m then Go)

Mode 4 — Scenario(3) & Runway (Sensitivity/Constraints)

• Base/Upside/Downside: ROI/Payback sensitivity
• Runway/Capacity constraints reflected in Sequencing by Quarter (Q-by-Q start order/parallel limits)

Mode 5 — Pre-mortem, Kill-Switch & Adaptability (Extended)

• Pre-mortem: Top-5 failure causes
• Leading Indicators (3): metrics/queries/logs + measurement frequency
• Kill-Switch Criteria: metric + period(window) + Action (Pause/Pivot/Stop)
• Pivot Triggers: transition conditions
• Market Triggers (new): external signals (competitor launch/policy/platform change/cost curve/ecosystem events)
• Adaptation Plan (new): discussion and priority reallocation rules (Playbook) when signals detected

Mode 6 — Stakeholder Lens (Extended)

• Investor/Board Lens: North Star/OKR/Unit Economics ↔ IR Narrative fit, Volatility/Seasonality risks
• Engineering Lens: Technical Debt/platform constraints, team autonomy/motivation
• GTM Lens: Sales/Marketing Narrative Fit, channel/sales motion alignment
• Customer Support Lens: expected issue volume/type, Help Center/Macro/Tooling preparation

------------------------------

Evidence Gate (Auto)

------------------------------

Core (≥4 required for Proceed):

1. Last-180d Signals ≥ 3 (founder/customer notes, 3–5 conv. summaries, tickets/reviews, competitor moves)
2. "Riskiest" assumptions ≥ 2 (tagged/listed)
3. Success criterion (OMTM or qualitative heuristic) exists
4. 1–2 week Discovery Plan or OKR draft
5. Versioned Evidence Hub link 1+ Advanced (bonus points):

• Baseline metric
• NSM/Metric tree or Funnel
• Experiment plan/results (hypothesis–method–outcome)
• TAM/SAM/SOM/ICP estimates Auto-decision:
• HOLD: Core ≤ 3 → Return 1-Day Evidence Sprint proposal
• PROCEED: Core ≥ 4 (must include #2, #3)
• AUTO-ESCALATE: PROCEED + Advanced ≥ 2
• PROVISIONAL: major blocker conditional (missing condition/deadline → error)

------------------------------

Scoring (0–5) & Anchors (Reference Points)

------------------------------

General Anchors: 0=no evidence, 1=memo level, 2=fragmentary/early, 3=solid/recent, 4=strong (metrics/experiment connected), 5=exemplary (reproducible)

P — Problem Diagnosis:

• 4: Recent 90-day behavioral data + JTBD mapping + broken assumptions list (evidence links) R — Reframe Opportunity:
• 4: Segment×Use-case clarity, competitive/substitute changes & timing logic quantified I — Intentional Bets:
• 4: Trade-off/Non-goals articulated + thresholds connected to OKR/NSM S — Systemized Execution:
• 4: Strategy→OKR/NSM→Discovery loop→Backlog/operating rhythm, calendar/issue tracker links M — Momentum & Meta-Reflection:
• 4: Retrospective/stop/resource reallocation history + learning summary (last 2 quarters)

Overall Score = average (equal weighting). HOLD prohibits scoring. Missing Stop-Rule allows Decision=Conditional only. Options < 3 triggers warning, < 2 triggers HOLD.

------------------------------

Outputs

------------------------------

1. review.md (≤900 words)

• Executive Summary (3–5): Decision (Approve/Conditional/Park/Kill), Top Strengths, Top Risks
• Impact × Confidence 2×2 (Options visual summary; Markdown table)
• PRISM Scores (0–5, Anchors applied) + evidence citations
• Pre-mortem/Kill-Switch, Market Triggers & Adaptation Plan
• Next Review Date

2. review.json (machine-readable; example schema)

{
  "title": "PRISM Strategy Review — <Area> — <YYYY-MM-DD>",
  "eligibility": "PROCEED|HOLD|AUTO-ESCALATE|PROVISIONAL",
  "window": {"start":"YYYY-MM-DD","end":"YYYY-MM-DD"},
  "doc": "<link>", "evidenceHub": "<link>", "owner": "@name",
  "principle_alignment": {
    "mission_vision_coherence": "...",
    "vision_narrative_1_3y": "...",
    "value_thesis": "..."
  },
  "scores": {"P":4,"R":3,"I":4,"S":3,"M":4,"overall":3.6},
  "options": [{
    "name":"Do Nothing",
    "assumptions":[{"text":"...","confidence":0.4},{"text":"...","confidence":0.6}],
    "investment":{"fte":2,"capex":0,"opex":50000,"tti_weeks":4},
    "impact":{"nsm_delta":"~0–1%","okr":["KR1","KR2"]},
    "risks":["...","...","..."],
    "non_goals":["..."],
    "decision_rule":"..."
  }],
  "capital_allocation":{"total_fte": "...", "budget":"..."},
  "scenarios":{"base":{"roi":1.8,"payback_months":10},"upside":{},"downside":{}},
  "assumptions_tests":[
    {"assumption":"...", "test_method":"A/B|Survey|Logs|Causal",
     "threshold":"X%", "sample":"n", "duration_days":14,
     "confidence_before":0.3, "confidence_after": null,
     "stop_rule":"if < X% for 2w then pause"}
  ],
  "leading_indicators":[{"name":"...","query_or_log":"...","cadence":"weekly"}],
  "kill_switch":[{"metric":"D+7 Retention","threshold":"<X%","window_days":14,"action":"Pause"}],
  "market_triggers":[{"signal":"Competitor launches X","monitor":"rss/source","play":"Re-prioritize Y→Z"}],
  "stakeholder_lens":{"investor_board":"...","engineering":"...","gtm":"...","customer_support":"..."},
  "next_steps":{"immediate":["..."],"nextReviewDate":"YYYY-MM-DD"},
  "citations":["[[ev-12|...]]","[[ev-13|...]]"]
}

------------------------------

Operational Rules

------------------------------

• Missing required input → HOLD + 1-Day Evidence Sprint (specify required items)
• Never hallucinate evidence; only cite materials within globs and inputs
• Generation order: review.json → review.md
• Action Items follow guides/product/process-task-list.mdc rules
• Token/length: md ≤ 900 words; Exec Summary 3–5 bullets

How to Use (Fastest Path)

1. Run Evidence Readiness auto-gate below. If status = Hold, run 1-day evidence sprint, then re-run gate. Do not score while on Hold.
2. If status = Proceed, score each PRISM dimension 0–5 using rubrics. Capture concrete evidence citations.
3. Record strengths, risks, decisions, and improvements using Output Template.
4. Compute Overall Score and decision. Assign owners and due dates for improvements.

Evidence Readiness Check (gate)

Track selection not required. Checklist below auto-applies Light or Standard interpretation based on evidence provided.

How to use (fastest path):

• Paste strategy doc link. Checklist will guide auto-judgment.
• Optionally add this mini-header at top of doc:
  • Doc:
  • Owner: @name
  • Window: YYYY-MM-DD ~ YYYY-MM-DD

Unified Evidence Checklist (auto-applied)

• Core items

  • 1–3 recent signals (≤ 180 days): founder/customer notes, 3–5 conversation summaries, support tickets/reviews, competitor moves
  • Top 2–3 riskiest assumptions clearly tagged ("Riskiest")
  • One success criterion: OMTM or qualitative success heuristic
  • 1–2 week discovery plan or OKR draft
  • One consolidated evidence hub link (doc/sheet/folder)

• Advanced items

  • Baseline metric (rough logs or estimates acceptable)
  • North Star/metric tree or lightweight funnel
  • Experiment plan/results with hypothesis-method-outcome
  • User/market size estimate (rough TAM/SAM/SOM or ICP count)

Auto-decision rules

• Proceed (eligible to score PRISM): ≥ 3 Core items present
• Hold: ≤ 2 Core items → run 1-day evidence sprint, then re-check
• Auto-escalate to Standard interpretation for future reviews if either:
  • Core met + ≥ 2 Advanced items, or
  • Both baseline metric and plan/OKR exist together

Operating notes

• If quantitative data missing, qualitative/market signals count toward Core
• Any tool fine; keep links consolidated in one place
• If on Hold, 1-day sprint must be time-boxed (e.g., 3 interviews summary; capture 10 competitive leads)

Scoring Overview

• Eligibility: Only compute and record PRISM scores if Evidence Readiness status = Proceed.
• Scale per dimension: 0–5
  • 0: Not present; 1: Minimal; 2: Emerging; 3: Solid; 4: Strong; 5: Exemplary
• Weighting: Equal weighting across PRISM (20% each)
• Overall Score = average of P, R, I, S, M

---

P — Problem Diagnosis

What good looks like:

• Identifies core challenge, distinguishes causes vs. symptoms, evidence-based.
• Articulates what's changing (market/tech/user behavior) and which assumptions broke.
• Ties user Jobs-To-Be-Done and unmet needs to business context.

Ask:

• What's changing in market, technology, or user behavior?
• Which assumptions are no longer holding true?
• What are users really trying to get done, and what blocks them?

Evidence to cite:

• Market/tech trend brief, user research synthesis, JTBD statements, opportunity map, analytics showing friction.

Red flags:

• Problem framed as feature gap, not causal diagnosis
• No user or market evidence; dated insights
• Confusing symptom metrics with root cause

Rubric (0–5):

• 0–1: Vague or opinion-only problem; no causal chain
• 2: Some evidence; symptoms and causes mixed
• 3: Clear causal story with recent evidence; scope reasonable
• 4: Strong causal model, validated with multiple sources and segmentation
• 5: Robust problem thesis with leading/lagging indicators and falsification paths

Reviewer outputs:

• Summary of diagnosed core problem(s)
• Top 1–2 causal insights supported by evidence links
• Improvement suggestions (e.g., missing segments, contradictory signals to reconcile)

---

R — Reframe Opportunity

What good looks like:

• Converts problem into compelling strategic opportunity with timing rationale.
• Clarifies what would shift for users/business if solved well; beyond incrementalism.

Ask:

• Why is now the right time?
• If solved, what materially changes for users and the business?
• What makes this more than incremental?

Evidence to cite:

• TAM/SAM/SOM or value pool, competitive whitespace, timing catalysts (tech, regulation, distribution), user willingness signals.

Red flags:

• Jumps to features; lacks clear opportunity thesis
• No timing rationale; ignores competitive dynamics
• Overpromises without mechanism of change

Rubric (0–5):

• 0–1: No opportunity framing
• 2: Opportunity stated but generic; weak timing
• 3: Clear opportunity thesis with timing and scope
• 4: Differentiated framing with credible mechanism of advantage
• 5: Compelling, time-sensitive opportunity with defensibility and user/business shift defined

Reviewer outputs:

• Opportunity statement and timing logic
• Key differentiators and value mechanism
• Improvements (e.g., clarify defensibility, quantify upside bounds)

---

I — Intentional Bets

What good looks like:

• Bold, testable strategic choices with explicit trade-offs and non-goals.
• Crisp Where-to-Play / How-to-Win statements linked to measurable hypotheses.

Ask:

• What are the few big bets? What are we not doing?
• How are bets testable within timeframe with clear success thresholds?
• How do bets connect to hypotheses and upside?

Evidence to cite:

• Bet statements, non-goals, hypotheses with metrics, decision logs, alternative paths considered.

Red flags:

• Feature lists disguised as strategy; no trade-offs
• No hypotheses or thresholds; can't be falsified
• Everything is a priority

Rubric (0–5):

• 0–1: No real choices; themes only
• 2: Choices exist but lack trade-offs or tests
• 3: Defined bets with non-goals and baseline hypotheses
• 4: Testable, time-bound bets with success/failure thresholds and resourcing
• 5: Coherent portfolio of bets with staged options and kill/sustain criteria

Reviewer outputs:

• List of bets and non-goals with links
• Testability assessment and gaps
• Improvements (e.g., define thresholds, add non-goals, narrow scope)

---

S — Systemized Execution

What good looks like:

• Strategy translates into OKRs, North Star/inputs, and discovery loops.
• Bets shape backlog, governance, and learning checkpoints.

Ask:

• Do OKRs reflect the bets vs. BAU metrics?
• Are discovery/validation loops defined with leading indicators?
• Are check-ins and operating rhythms in place?

Evidence to cite:

• OKRs, North Star metric tree, discovery plan, experiment backlog, rituals calendar, resource plan.

Red flags:

• OKRs disconnected from bets; vanity or lag-only metrics
• No discovery cadence; roadmap-first planning
• Lack of ownership or governance

Rubric (0–5):

• 0–1: No bridge to execution
• 2: Some metrics or OKRs, weak alignment
• 3: OKRs and discovery plan map to bets
• 4: Robust metric tree, clear owners, and learning cadence
• 5: Tight strategy-to-execution system with adaptive re-planning and leading→lagging linkage

Reviewer outputs:

• Execution readiness assessment (OKRs, metrics, loops, owners)
• Risks to delivery or learning velocity
• Improvements (e.g., define leading indicators, set decision checkpoints)

---

M — Momentum & Meta-Reflection

What good looks like:

• Built-in reflection to avoid strategic drift; deliberate learning objectives.
• Regularly says no; challenges assumptions; documents learning.

Ask:

• What did we learn this quarter? What changed?
• Where did we say yes when we should've said no?
• What are we not challenging enough?

Evidence to cite:

• Quarterly retro, decision reviews, stopped work examples, updated assumptions log.

Red flags:

• No retros; strategy as static doc
• Accretion of scope; no stopped initiatives
• Assumptions never revisited

Rubric (0–5):

• 0–1: No reflection mechanisms
• 2: Ad-hoc reflection; little impact on choices
• 3: Quarterly reviews with documented learning and some course-correction
• 4: Systematic reflection tied to decisions and resource shifts
• 5: Strong learning culture with explicit un-learning and portfolio rebalancing

Reviewer outputs:

• Reflection mechanisms and learning culture assessment
• Improvements (e.g., add stop criteria, schedule decision reviews)

---

Output Template (copy for each review)

Title: PRISM Strategy Review — <Product/Area> — <YYYY-MM-DD>

Eligibility (auto-gate): Proceed / Hold / Auto-escalated
Doc: <link>
Evidence Hub: <link>
Owner: @name
Window: YYYY-MM-DD ~ YYYY-MM-DD

Executive Summary (3–5 bullets)
- Key call: Approve / Provisionally Approve (with conditions) / Revise & Resubmit
- Top strengths
- Top risks/assumptions to resolve

Scores (0–5)
- P – Problem Diagnosis: <score> — evidence: <links>
- R – Reframe Opportunity: <score> — evidence: <links>
- I – Intentional Bets: <score> — evidence: <links>
- S – Systemized Execution: <score> — evidence: <links>
- M – Momentum & Meta-Reflection: <score> — evidence: <links>
- Overall Score: <avg>

Findings & Rationale
- P: <summary of diagnosis quality>
- R: <summary of opportunity framing>
- I: <summary of bets and trade-offs>
- S: <summary of strategy→execution alignment>
- M: <summary of reflection mechanisms>

Priority Improvements (numbered, with owners & dates)
1) [P/R/I/S/M] <improvement>, Owner: <name>, Due: <date>, Success: <measure>
2) ...

Risks & Assumptions to Validate
- Assumption: <text> — Test: <method> — Threshold: <value> — By: <date>

Decision & Next Steps
- Decision: <approve status>
- Immediate actions: <bullets>
- Next review date: <date>

Reviewer Checklist Before Submission

• Evidence Readiness status recorded (Proceed/Hold/Auto-escalated)
• If Hold, 1-day evidence sprint was executed and gate re-run
• Links consolidated to single evidence hub
• All scores justified with links to evidence
• Improvements are actionable, prioritized, and owned
• Decision is explicit; conditions (if any) are measurable
• Next review date is set; learning objectives defined

Common Anti-patterns (Quick Reference)

• Strategy as feature list; no explicit trade-offs or non-goals
• Lagging metrics only; no leading indicators for discovery
• Untested assumptions presented as facts; no falsification plan
• Over-expansive scope; unclear target segment or use case
• Reflection is performative; no stopped work or resource reallocation