Purpose
Turn an unstructured description of a data estate into a scored, tiered inventory — rating every source on five decision dimensions and producing a prioritization matrix with ranked recommendations for where to invest, monetize, remediate, or retire. Solves the problem of a data organization that owns dozens of assets but has no defensible way to say which ones are strategically worth funding.
When to Use
Invoke this skill when the user asks to:
- Catalog or inventory available data sources and rate their value
- Assess which data assets are worth investing in, cleaning up, or monetizing
- Prepare a data-estate assessment as input to a data strategy or roadmap
- Identify unique or defensible data the organization holds versus commodity data
- Compare data sources on quality, coverage, or strategic fit before prioritizing spend
- Build a recurring, repeatable data asset review for governance or planning cycles
Inputs Required
Before starting, collect:
- Data sources list: The assets to assess — name and 1–2 sentence description of each (systems, tables, feeds, warehouses, third-party datasets)
- Business context: What the organization does and the strategic goals the data should serve
- Intended use cases: What decisions, products, or initiatives these assets are meant to feed (analytics, ML, monetization, reporting)
- Known quality/access issues: Any documented problems — gaps, latency, PII constraints, ownership ambiguity
- Refresh characteristics: How often each source updates, if known (real-time, daily, batch, static)
- Scope of assessment: Full estate vs. a specific domain, and whether the goal is prioritization, monetization, or remediation
- Comparables available: Whether similar data can be bought or built externally (for uniqueness scoring)
If refresh cadence or comparables are unknown, score conservatively on those dimensions and flag the assumption explicitly.
Instructions
Step 1: Normalize the Asset List
Restate each described source as a discrete inventory line with a consistent short name, a one-line description, and its apparent primary domain (customer, transactional, operational, external, derived). Merge duplicates and split any "source" that is actually several distinct assets. Confirm the count before scoring.
Step 2: Score Each Asset on Five Dimensions (1–5)
Rate every asset on each dimension using the anchored scale below. Never assign a score without a one-clause justification.
Uniqueness — how defensible / hard to replicate
- 5: Proprietary, cannot be bought or rebuilt elsewhere
- 3: Partially unique; some external substitute exists
- 1: Commodity data freely available or purchasable
Coverage — breadth and completeness relative to the use case
- 5: Comprehensive across entities, geographies, and time
- 3: Meaningful but partial coverage with known blind spots
- 1: Sparse, fragmentary, or narrow slice only
Quality — accuracy, consistency, structure, trustworthiness
- 5: Clean, governed, validated, low error rate
- 3: Usable but requires meaningful cleaning/reconciliation
- 1: Unreliable, inconsistent, or poorly documented
Timeliness — freshness relative to what the use case needs
- 5: Real-time or fresh enough that latency is irrelevant
- 3: Periodic refresh acceptable for most uses
- 1: Stale, static, or latency undermines value
Strategic Relevance — fit to stated goals and use cases
- 5: Directly enables a priority strategic initiative
- 3: Supports secondary initiatives or general analytics
- 1: Tangential to current strategy
Step 3: Compute Composite Score and Tier
For each asset, sum the five scores (max 25) and assign a tier:
- Tier 1 — Crown Jewels (20–25): Defensible, high-relevance assets to protect and leverage
- Tier 2 — Strategic Assets (15–19): Valuable, worth targeted investment to elevate
- Tier 3 — Utility Assets (10–14): Operationally useful but not differentiating
- Tier 4 — Liabilities/Candidates to Retire (below 10): Low value, high maintenance, or redundant
Weight Strategic Relevance and Uniqueness more heavily in tie-breaks — a high-quality commodity asset is worth less strategically than a moderately clean proprietary one.
Step 4: Build the Value-Effort Prioritization Matrix
Plot each asset on two axes: Strategic Value (composite score) vs. Effort to Unlock (inferred from Quality + Timeliness gaps). Assign each asset to one quadrant:
- Leverage Now: High value, low effort — activate immediately
- Invest: High value, high effort — fund remediation to unlock
- Maintain: Lower value, low effort — keep running, don't over-invest
- Reconsider: Lower value, high effort — retire, consolidate, or defer
Step 5: Gap and Opportunity Analysis
Identify what the estate is missing relative to the stated use cases:
- Coverage gaps where no owned asset serves a priority initiative
- Assets whose value is trapped behind a single fixable dimension (e.g., strong data with only a timeliness problem)
- Redundant or overlapping sources that should be consolidated
- Monetization candidates: high-uniqueness, high-coverage assets with external demand potential
Step 6: Ranked Recommendations
Produce 6–10 recommendations ordered by expected strategic impact. Each recommendation states: the asset(s) involved, the action (leverage, invest, remediate, consolidate, monetize, retire), the dimension it moves, and the outcome it enables. Be decisive — name specific assets, not general principles.
Output Format
- Structure: Normalized asset list → scored inventory table → tier summary → prioritization matrix → gap analysis → ranked recommendations
- Length: 1,000–1,400 words plus tables
- Format: A scored table (one row per asset, five dimension columns, composite, tier) plus structured prose for matrix, gaps, and recommendations
- Sections: Inventory Overview → Scored Asset Table → Tier Breakdown → Value-Effort Matrix → Gaps & Opportunities → Ranked Recommendations
- Tone: Data strategy advisor to an executive — analytical, decisive, investment-minded; every score defended, no hedging on tiers
Quality Checklist
Before delivering output, verify:
- [ ] Every asset from the input appears exactly once in the scored table (count matches Step 1)
- [ ] Each asset has all five dimension scores plus a composite and a tier
- [ ] No dimension score appears without a one-clause justification
- [ ] Every asset is placed in exactly one matrix quadrant
- [ ] Gap analysis names at least one coverage gap or trapped-value asset (not "no gaps found" unless justified)
- [ ] Recommendations are ranked by impact and each names specific assets and a specific action
Notes
- The most common finding is a high-quality asset trapped behind one weak dimension — timeliness or access — where a modest investment moves it up a full tier. Surface these first; they are the cheapest wins.
- Uniqueness is the dimension most often overrated by asset owners. Pressure-test any 5 against whether a competitor could buy or rebuild the equivalent.
- When run recurringly, keep dimension anchors constant across cycles so tier movement is comparable over time — that trend is itself a strategy signal.
- If the estate is large, offer to run a first pass at the domain level before scoring individual assets, so scoring effort concentrates on the domains that matter.
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