meta-living-plan-governance

name	meta-living-plan-governance
description	Use when a business plan is being finalised for submission or internal use, and the engine must specify how the plan will be maintained. Use the relevant plan-section skill for section drafting.
metadata	{"portable":true,"compatible_with":["claude-code","codex"]}

Living Plan Governance Skill

Overview

A business plan is not a one-off artefact for a funding committee. It is a living document that should evolve with the business across its entire lifetime. This skill installs the operating cadence by which the plan stays alive: what metrics feed it, how often each section is reviewed, who owns each section, where decisions get logged, what variance triggers re-planning, and when content is archived.

This skill is mandatory for SaaS / ICT plans because their assumptions decay faster than traditional industries: pricing experiments resolve in weeks, channel performance shifts quarterly, AI cost economics change every 6 months, and competitor moves can invalidate strategy overnight.

Use When

• A business plan is being finalised for submission or internal use, and the engine must specify how the plan will be maintained
• A funded company is setting up its operating rhythm post-investment
• An existing company has a static plan that has not been updated in 6+ months and is now misaligned with reality
• An investor or board is asking how the plan-vs-actual variance will be governed
• A digital-first or SaaS-first business needs explicit cadences because its market changes faster than annual planning can absorb

Do Not Use When

• The plan is a one-off proposal document for an external party (use proposal-architect instead)
• The business is in deep crisis where survival decisions override planning rhythm (stabilise first)
• The user wants a generic project plan — this is specifically the business-plan-as-living-document discipline

Required Inputs

Input	Source / provider	Required?	If absent
Living Plan Governance brief and decision audience	Client, plan owner, or approved project files	Yes	Stop before making a recommendation; state the missing decision context.
Claims, assumptions, and supporting evidence	Source register, model, research notes, interviews, or operating records	Yes	Separate known facts from assumptions and return a qualified gap list.
Authority and delivery constraints	Requesting owner and repository instructions	Yes	Remain read-only and produce a draft or review only.

• The finalised or near-finalised business plan
• The list of plan assumptions (from each section) that are material to outcomes
• The current team and proposed roles (to assign ownership)
• The KPI dashboard or metrics available (or the gap, if nothing exists)
• The board / investor cadence expectations
• The country / regulatory reporting calendar (annual returns, tax filings, regulatory submissions)

Workflow

1. Inventory the plan's assumptions — for each section (00 through 16), extract the load-bearing assumptions. For SaaS plans, these typically include: ARR growth, churn, NRR, CAC, conversion rates, gross margin, payment-rail availability, FX rate, AI cost per tenant, key-customer concentration, regulatory status.
2. Map each assumption to a metric / signal — every assumption needs a measurable proxy. If a metric does not yet exist, note the instrumentation requirement.
3. Assign cadence to each section — weekly / monthly / quarterly / annual / trigger-only. See the cadence table below.
4. Assign owner to each section — CEO / CFO / Head of Product / Head of GTM / Head of CS / Founder / Board chair. Avoid "everyone owns it" — that means no one does.
5. Define variance thresholds — at what plan-vs-actual gap does the section trigger a re-plan? Default: ±15% on revenue metrics, ±25% on cost metrics, ±5pp on churn, ±10pp on NRR.
6. Set up the decision log — a single canonical location (Notion / Coda / Confluence / shared doc) where every material decision is recorded with: date, decision, alternatives considered, evidence, decision-maker, expected outcome, review date.
7. Define trigger-replan events — explicit events that force an immediate re-plan regardless of cadence: founder departure, key-customer churn, regulatory shock, FX shock >10%, technology shock (AI-cost spike, platform deprecation), funding round closure, major hire / fire.
8. Define sunset policy — when does a section's content get archived? Default: when the section's assumptions have been replaced by ≥2 quarterly updates, and the original content no longer reflects current strategy.
9. Configure the Mission/Strategy/Projects/Omissions/Tracking (MSPOT) artefact — adopt HubSpot's MSPOT (Cotton) as the annual artefact that summarises the living plan. Mission rarely changes; Strategy changes annually; Projects are the 4–5 big initiatives; Omissions are the explicit "not this year" list; Tracking is the KPI dashboard.
10. Document the cadence calendar — produce a 12-month rolling cadence calendar (weekly stand-up, monthly business review, quarterly board pack, annual strategy refresh).

Decision, stop, and recovery controls

• Decision point: confirm that the requested output is the living-plan governance calendar and that the decision concerns which variance triggers correction, reforecast, or strategy review.
• Stop condition: halt the affected conclusion if required evidence is missing (metric owners, plan baseline, decision rights, and review cadence) or if the work could lead to this identified risk: allowing a stale plan to govern current spending.
• Recovery: obtain the missing record or reviewer, repeat the affected check, and update the exception record before release.

Cadence Table by Section

Section	Weekly	Monthly	Quarterly	Annual	Trigger
01 Exec Summary				refresh	acquisition / pivot
02 Company Overview				refresh	legal change
03 Products		KPI review	roadmap re-plan		major launch / sunset
04 Market			competitor scan	TAM refresh	major entrant / regulation
05 Target Market			ICP refresh	persona refresh	win/loss data shift
06 Competitive			scan	full re-analysis	major M&A
07 Marketing/Sales	metrics	MBR	full strategy review	full re-plan	channel collapse / 2× growth
08 Operations		SLA review	process audit	annual review	outage / SLA breach
09 Management/Team		hiring update	org review	annual offsite	senior departure
10 Financial Projections	actuals vs plan	full review	re-forecast	new 3yr plan	variance >15%
11 Funding Request		runway check			round opens / closes
12 Risk		risk-flag review	full re-analysis		new top-3 risk
13 Implementation	sprint planning	milestone review	gate review		milestone slip >30 days
14 AI Integration	tool stack		full review		new model / cost shock
16 Sustainability			impact KPI	full review	regulatory change

AI Living-Plan Cadence (mandatory for AI-feature-led plans)

When AI is material to the plan (typically >2% of ARR or load-bearing to product), add the following AI-specific cadence on top of the section-by-section cadence above. Each element has explicit owner and variance threshold.

AI element	Cadence	Owner	Variance threshold	Trigger-replan condition
Eval suite scores (accuracy / hallucination / refusal / latency)	weekly	Head of AI	-5pp on any metric	-10pp in single week
Per-tenant AI cost (median + top decile)	monthly	CFO + Head of AI	+20% MoM	sustained +30% over 2 months
AI Gross Margin	monthly	CFO	-3pp MoM	-5pp QoQ
AI-cost-as-%-of-ARR	monthly	CFO	>15% alarm	>20% absolute
Hallucination rate (production sampling)	monthly	Head of AI + Head of QA	+1pp absolute	sev-1 hallucination event = immediate
Cache-hit ratio	weekly	Head of AI	-10pp from baseline	-20pp
Token usage per query	weekly	Head of AI	+25% WoW	+50% WoW
Retraining trigger watch	monthly	Head of AI	drift >threshold	concept drift confirmed
Model-deprecation watch	monthly	Head of AI / CTO	provider notice	deprecation announced
Provider pricing watch	weekly + per-change	Head of AI / CTO	any change	major change (>30%)
AI moat reassessment	quarterly	CEO + Head of Strategy	-3 points on rubric	competitor parity claim
Regulatory AI watch	quarterly	Head of Legal / Compliance	new rule	active enforcement against analogue
AI-incident log	continuous + monthly review	Head of AI	any sev-1	sev-1 customer-impact event
AI vendor concentration	quarterly	CFO + CTO	>80% on single provider	provider EULA / pricing shock
AI talent retention	quarterly	Head of People + CTO	>20% AI-team attrition	departure of AI lead

The AI cadence is operationalised through:

• skills/pipeline/10-financial-projections/saas-ai-unit-economics-and-cogs/SKILL.md — economics + cost
• skills/pipeline/10-financial-projections/saas-ai-cost-of-tenant-calculator/SKILL.md — per-tenant cost
• skills/pipeline/12-risk-analysis/saas-ai-risk-and-stress-test/SKILL.md — risk register + stress
• skills/pipeline/06-competitive-analysis/saas-ai-moat-and-defensibility/SKILL.md — moat reassessment
• skills/meta-finance/meta-ai-bankability-and-investor-readiness/SKILL.md — scorecard refresh
• skills/pipeline/08-operations-plan/references/ai-cost-and-vendor-management-ops.md — operations
• skills/pipeline/16-sustainability-strategy/saas-ai-sustainability-and-ethics/SKILL.md — ethics + sustainability
• skills/meta-reporting/meta-board-and-investor-reporting/references/saas-ai-board-pack-section.md — reporting

Agent Living-Plan Cadence (mandatory when the plan ships an agent or multi-agent product)

When the plan is an agent business, add the Agent Cadence layer on top of the AI cadence above. The agent cadence is specified in:

references/agent-cadence-table.md

It covers weekly (task success, intervention rate, cost-per-resolved, tool reliability, branch / loop breaches, eval scores, cache, token usage, provider pricing scan, prompt-injection scan), monthly (agent GM, HITL / tool / retry cost shares, agent revenue attribution, irreversibility / autonomy incidents, audit / kill-switch / red-team drills, eval coverage, model-deprecation watch, contestability, channel reliability, AI Safety Lead retention, sustainability KPIs, misuse detection), quarterly (moat reassessment, autonomy expansion review, regulator watch, stress refresh, tabletop exercise, reserves review, bankability rescore, valuation refresh, foundation-model platform risk, comparable scan, insurance review, tool registry proprietary share, AI Safety Lead succession, talent retention, customer rollout, training-data provenance audit, jobs-impact tracking), annual (full simulation, external ethics review, Wardley refresh, comp scan, sustainability annual review, multi-country expansion review).

The agent cadence is operationalised through:

• skills/pipeline/10-financial-projections/saas-agent-unit-economics-and-cogs/SKILL.md — cost per resolved
• skills/pipeline/07-marketing-sales-strategy/saas-agent-pricing-strategy/SKILL.md — pricing
• skills/pipeline/06-competitive-analysis/saas-agent-moat-and-wrapper-risk/SKILL.md — moat
• skills/pipeline/12-risk-analysis/saas-agent-risk-and-stress-test/SKILL.md — risk + stress + drills
• skills/pipeline/09-management-team/saas-agent-talent-strategy/SKILL.md — talent
• skills/pipeline/03-products-services/saas-agent-product-strategy-and-roadmap/SKILL.md — autonomy ladder
• skills/pipeline/13-implementation-timeline/saas-agent-implementation-timeline/SKILL.md — rollout
• skills/pipeline/16-sustainability-strategy/saas-agent-sustainability-and-ethics/SKILL.md — ethics + sustainability
• skills/meta-finance/meta-agent-bankability-and-investor-readiness/SKILL.md — scorecard
• skills/meta-finance/meta-agent-valuation-adjustments/SKILL.md — valuation
• skills/meta-reporting/meta-agent-board-and-investor-reporting/SKILL.md — reporting
• skills/pipeline/08-operations-plan/references/agent-operations-runbook-summary.md — operations runbook

The MSPOT Annual Artefact

Cotton (HubSpot) MSPOT structure:

• Mission — the unchanging "why we exist" (refreshed only with company-defining events)
• Strategy — the "how we win this year" (annual refresh; quarterly review)
• Projects — the 4–5 big initiatives this year; everything else is BAU or omitted
• Omissions — the explicit list of things we are NOT funding this year (the "no" list)
• Tracking — the KPI dashboard with explicit targets and review cadence

The Omissions section is the highest-leverage discipline; it's how the organisation limits the CEO's appetite and prevents indigestion.

The Decision Log Standard

Every material decision recorded with:

• Date
• Decision title and one-line description
• Alternatives considered (do-nothing always one of them — Haines discipline)
• Evidence considered (data, customer interviews, market signals)
• Assumptions made
• Decision-maker(s)
• Expected outcome / success criteria
• Review date (when do we check if this worked)
• Outcome (filled in at review date — Haines post-implementation audit discipline)

The Plan-vs-Actual Variance Protocol

When variance exceeds threshold:

1. Diagnose — is the variance from execution failure (we missed the plan) or from environmental shift (plan assumption was wrong)?
2. Reconcile — does this variance invalidate downstream sections? (A churn miss propagates to LTV → unit economics → runway → funding ask.)
3. Re-plan — produce a delta to the affected sections; do not rewrite untouched sections.
4. Log — the variance, the diagnosis, the change, the decision-maker in the decision log.
5. Communicate — notify the board / investors with the diagnosis and remediation; do not surprise them at the next board meeting.

Trigger-Replan Events (force immediate re-plan regardless of cadence)

• Founder or C-suite departure
• Loss of customer representing >10% of ARR
• Currency depreciation >10% in 30 days
• Regulatory event invalidating a material assumption (e.g. payment-rail ban, data-residency law)
• Technology shock (key API deprecated, AI cost shift, security breach)
• Funding round closure (new capital changes the runway plan)
• M&A activity (own acquisition, key competitor acquired)
• Major key-supplier failure (AWS region outage extending >24h, key payment-gateway suspension)

Quality Bar

• Every load-bearing assumption in the plan has an owner, a metric, and a cadence
• Every section has a defined review cadence
• A canonical decision log exists and is being maintained
• Variance thresholds are explicit, not implicit
• Trigger-replan events are enumerated, not vague
• MSPOT (or equivalent) is produced annually and visible to the team
• Plans more than 6 months old without quarterly updates should be flagged as stale

Anti-Patterns

• Treating the business plan as a fundraising artefact that gets archived after the round closes

• "We'll update it when we have time" — no, you won't; install the cadence

• KPI dashboards that are not tied to plan assumptions

• Decisions made informally without logging — institutional memory dies with founder turnover

• Sections that haven't been touched in 12+ months but are still treated as canonical

• Mission / Strategy / Vision in one document with no Omissions list

• Applying the wrong neighbouring route to meta living plan governance. Correction: confirm the decision and route to the named neighbour before analysis.

• Treating an assumption as verified evidence. Correction: label it, cite its source or owner, and assign a verification action.

• Recommending action without a decision threshold. Correction: state the measurable acceptance condition and review trigger.

• Recording an unavailable check as passed. Correction: mark it not assessed and state the consequence for the decision.

• Mutating or publishing during an analysis-only task. Correction: remain read-only until the owner gives explicit authority.

Outputs

Artefact	Consumer	Observable acceptance condition
Living Plan Governance deliverable	Named decision-maker or plan author	The recommended choice, assumptions, countercase, and next action are explicit.
Evidence and exception register	Reviewer, funder, board, or implementation owner	Every load-bearing claim is sourced or labelled as an assumption; missing checks are not shown as passes.

• A cadence calendar for the next 12 months
• A per-section ownership and variance-threshold matrix
• A populated decision-log template
• An MSPOT artefact for the current year
• A trigger-replan event list customised to the business
• A handoff document to the operating team explaining how to maintain the plan

References

• references/living-business-plan-operating-model.md — full operating-model reference
• book-extractions/cotton-run-a-saas-business-extraction.md — MSPOT discipline (section 10)
• book-extractions/haines-how-to-create-a-business-case-extraction.md — post-implementation audit, decision-log discipline
• book-extractions/walling-saas-playbook-extraction.md — customer-conversation cadence as living-strategy input
• skills/meta-strategy/meta-quarterly-gameplan/SKILL.md — quarterly operational rhythm (sister skill)
• skills/meta-strategy/meta-monitoring-evaluation/SKILL.md — KPI dashboard discipline (sister skill)

Africa / Uganda Context

• Annual planning cycles often align with funding-cycle calendars (Q4 USAID / FCDO budgets, Q1 corporate budgets, Q3 multilateral grants). Sync the cadence calendar with the funder calendar.
• Quarterly board packs are often required by DFIs (UDB, AfDB, IFC) — adopt their format early to avoid re-work.
• Trigger-replan events in Africa include: FX shock (KES, UGX, NGN can move 10–30% in a quarter), payment-rail policy change (Central Bank circulars affecting M-Pesa, MoMo, fintech), regulatory shock (data-localisation laws, sector-specific licensing).
• Decision log should also record political / stakeholder context — many decisions in African enterprise are co-determined by political-economy factors that pure financial logic misses; the log preserves this knowledge.

Evidence Produced

Evidence	Format	Acceptance condition
Living-plan governance calendar decision trace	Sources, calculations, assumptions, countercase, and selected action	A reviewer can trace the selected action and rejected alternatives to the cited inputs.
Exception record	Failed and not-assessed checks with owner and due action	The register exposes every unresolved exception that could lead to allowing a stale plan to govern current spending.

Capability and Permission Boundaries

Read supplied records and use non-mutating checks to produce the living-plan governance calendar; updating the approved governance calendar and decision log is permitted when requested. Do not publish, contact third parties, alter live systems, commit funds, or claim legal, tax, audit, valuation, ESG, or investment assurance without the owner's explicit authorisation and the appropriate reviewer.

Degraded Mode

If metric owners, plan baseline, decision rights, and review cadence cannot be obtained, return a qualified living-plan governance calendar covering only the checks that remain supportable. Leave this decision unresolved: which variance triggers correction, reforecast, or strategy review. Record the evidence owner and next check; an inaccessible source, tool, or reviewer is never a pass.

Decision Rules

Decision condition	Action	Failure or risk avoided
Evidence is sufficient to decide: which variance triggers correction, reforecast, or strategy review	Record the conclusion, source trail, owner, and review trigger in the living-plan governance calendar.	Risk of allowing a stale plan to govern current spending
Material evidence conflicts or remains uncertain	Apply both candidate variance triggers to prior periods and choose the one that prompts action without creating routine false alarms.	Selecting an option without resolving the decision-relevant uncertainty
Required evidence is missing: metric owners, plan baseline, decision rights, and review cadence	Mark the decision on which variance triggers correction, reforecast, or strategy review not assessed in the living-plan governance calendar, and send it to the planning owner and chief executive.	Otherwise, the work risks allowing a stale plan to govern current spending

Quality Standards

Accept the living-plan governance calendar only when evidence is sufficient for this decision: which variance triggers correction, reforecast, or strategy review. Assumptions and countercases remain visible, calculations and cross-references reconcile, and the reviewer can see how the recommendation addresses the risk of allowing a stale plan to govern current spending.

Worked Example

A monthly revenue miss comes from delayed permits rather than demand. The governance process records the variance, owner, recovery date, and replan trigger instead of rewriting the annual target immediately.