Purpose
Score a specific finance or accounting process against a defined 1–5 maturity scale across people, process, data, and technology — producing a defensible per-dimension score, a weighted overall level, and a ranked set of next-level improvements. Replaces subjective "we're pretty manual" self-assessments with a structured diagnostic that supports transformation prioritization and executive buy-in.
When to Use
Invoke this skill when the user asks to:
- Assess the maturity of a finance process (month-end close, AP, AR, reconciliations, FP&A, treasury, expense management)
- Benchmark current-state process maturity before a systems or transformation project
- Justify investment in automation or a new finance tool with a maturity gap analysis
- Build a roadmap for advancing a process from its current stage to the next level
- Compare maturity across multiple finance processes to decide where to focus first
- Prepare a maturity summary for a controller, CFO, or steering committee
Inputs Required
Before starting, collect:
- Process being scored: The specific finance process (e.g., "month-end close," "3-way match AP," "bank reconciliation")
- Current-state description: How the process runs today — steps, handoffs, tools, and where manual effort concentrates
- Team and roles: Who performs the work, headcount, skill level, and reliance on key individuals
- Systems and tools in use: ERP, spreadsheets, RPA, close-management or reconciliation software, integrations
- Data situation: Where data lives, how it's entered, error/rework rate, and how reconciled it is across sources
- Key pain points: Known bottlenecks, cycle time, error frequency, audit findings
- Target ambition (optional): Desired maturity level or timeframe for improvement
If team detail or data specifics are missing, score conservatively based on what the current-state description implies and flag the assumption.
Instructions
Step 1: Establish the Maturity Scale
Score every dimension against this 5-level scale. Anchor each score to observable evidence in the current-state description — not aspiration.
- Level 1 — Manual/Fragmented: Ad-hoc, undocumented, spreadsheet-driven, person-dependent. High rework and firefighting.
- Level 2 — Defined/Reactive: Documented steps exist but execution is inconsistent. Some standardization, still heavily manual.
- Level 3 — Standardized/Controlled: Consistent, repeatable process with controls, checklists, and defined ownership. Partial automation.
- Level 4 — Integrated/Automated: Systems connected, automated workflows, exception-based handling, near-real-time data. Metrics tracked.
- Level 5 — Optimized/Intelligent: Continuous improvement, predictive/AI-assisted, self-monitoring controls, minimal human touch on routine work.
Step 2: Score the People Dimension
Evaluate skills, ownership, and dependency risk.
- Signals to assess: Clear role ownership vs. hero dependence; cross-training and coverage; skill level (data entry vs. analysis vs. process design); capacity spent on manual work vs. value-add.
- Anchor the score: Level 1 = one person "knows how it works"; Level 3 = documented roles and backup coverage; Level 5 = team focused on analysis and improvement, routine work offloaded.
- Assign a 1–5 score with a 2–3 sentence justification citing the current-state evidence.
Step 3: Score the Process Dimension
Evaluate documentation, standardization, controls, and cycle time.
- Signals to assess: Documented and followed procedures; standardization across periods/entities; embedded controls and approvals; cycle time and rework rate; exception handling.
- Anchor the score: Level 1 = ad-hoc, undocumented; Level 3 = checklists, defined controls, repeatable; Level 5 = exception-based, continuously refined, self-monitoring.
- Assign a 1–5 score with a 2–3 sentence justification.
Step 4: Score the Data Dimension
Evaluate data quality, integration, and reconciliation.
- Signals to assess: Single source of truth vs. scattered spreadsheets; manual re-keying vs. automated flow; error/exception rate; reconciliation timeliness; auditability and traceability.
- Anchor the score: Level 1 = data re-keyed across disconnected spreadsheets; Level 3 = defined sources with periodic reconciliation; Level 5 = integrated, validated, near-real-time, self-reconciling.
- Assign a 1–5 score with a 2–3 sentence justification.
Step 5: Score the Technology Dimension
Evaluate the tooling stack and its usage.
- Signals to assess: ERP utilization vs. spreadsheet workarounds; workflow automation; specialized tooling (reconciliation, close, AP automation); integration between systems; use of RPA/AI.
- Anchor the score: Level 1 = spreadsheets and email; Level 3 = ERP used with some automation; Level 5 = integrated stack with intelligent automation and monitoring.
- Assign a 1–5 score with a 2–3 sentence justification.
Step 6: Calculate the Overall Maturity Level
- Weight the four dimensions equally by default (25% each) unless the user specifies priorities.
- Report the weighted average to one decimal (e.g., 2.5) and map it to the nearest named level with a one-line interpretation.
- Identify the lowest-scoring dimension as the primary constraint — the maturity of a process is often gated by its weakest link.
Step 7: Recommend Next-Level Improvements
For each dimension, specify what it takes to advance exactly one level — not a leap to Level 5.
- Provide 2–3 concrete actions per dimension, written as specific moves (e.g., "Document the close checklist with owner and due-time per task"), not goals.
- Rank all recommendations across dimensions into a Top 5 priority list, ordered by impact-to-effort, leading with the primary constraint dimension.
- For each top action, note the dimension it advances and the maturity level it moves toward.
Output Format
- Structure: Scorecard table first, then per-dimension narrative, then overall level, then ranked improvements
- Length: 700–1,000 words
- Format: A summary scorecard table (Dimension | Score 1–5 | Level Name | One-line rationale), followed by structured narrative sections
- Sections: Maturity Scorecard (table) → Per-Dimension Assessment (People, Process, Data, Technology) → Overall Maturity Level → Primary Constraint → Next-Level Improvements by Dimension → Top 5 Priority Actions
- Tone: Controller-to-controller — precise, evidence-based, non-flattering. Calls out the weakest link plainly.
Quality Checklist
Before delivering output, verify:
- [ ] All four dimensions (People, Process, Data, Technology) receive a numeric 1–5 score
- [ ] Each score is justified with evidence drawn from the provided current-state description, not generic assumptions
- [ ] The overall weighted maturity level is calculated and mapped to a named level
- [ ] The lowest-scoring dimension is explicitly named as the primary constraint
- [ ] Every improvement recommendation advances a dimension exactly one level and is a concrete action, not a goal
- [ ] The Top 5 priority list is ranked by impact-to-effort, not listed arbitrarily
Notes
- A process is only as mature as its weakest dimension — a Level 4 technology stack undermined by Level 1 data quality is functionally a Level 1 process. Weight the primary constraint accordingly in recommendations.
- Resist score inflation. If the description says "mostly in spreadsheets with a monthly reconciliation," that is Level 2–3, not 4, regardless of how disciplined the team is.
- Advancing one level at a time is the realistic path; skipping levels usually fails because the underlying discipline (documentation, controls) hasn't been built.
- When multiple processes are scored in sequence, the primary-constraint pattern often repeats — flag any dimension that scores low across processes as a function-wide investment priority.
How to use this skill
This file is a Claude Code skill in Anthropic's standard format — markdown with YAML frontmatter at the top.
Installing in Claude Code:
~/.claude/skills/<folder-name>/SKILL.md— auto-loads in every session anywhere on your machine (recommended for skills you'll use across multiple projects)<your-project>/.claude/skills/<folder-name>/SKILL.md— auto-loads only when working in that project (recommended for project-specific skills)
Restart Claude Code or open a new session to register.
Using in Claude.ai, ChatGPT, Gemini, or other AI tools:
- Drag this file into a new chat, or copy and paste its contents
- Send your task or request as your next message — the AI will apply the skill where it fits
For Claude Code skill details: https://docs.anthropic.com/en/docs/claude-code/skills