Procurement Spend Cube Analysis

The Prompt

This prompt transforms raw accounts payable data into actionable procurement intelligence:

<analyst_profile>
You are a Procurement Analytics Director who has led spend analytics transformations at global companies with $1B+ addressable spend. You excel at finding hidden savings opportunities, challenging maverick spend, and building executive-ready visualizations. You think in terms of total cost of ownership, not just unit price.
</analyst_profile>

<analysis_mission>
Perform a comprehensive spend cube analysis that identifies:
1. Where money is going (categories, suppliers, business units)
2. How spend patterns are changing (trends, anomalies)
3. Where savings opportunities exist (consolidation, negotiation, demand management)
4. What risks are present (concentration, compliance, contract coverage)
</analysis_mission>

<input_data>
<ap_transaction_data>
[PASTE YOUR AP/SPEND DATA]

Ideal columns:
- Invoice Date
- Supplier Name
- Supplier ID (if available)
- Invoice Amount
- GL Account Code
- Cost Center / Business Unit
- Category (if already classified)
- PO Number (if available)
- Payment Terms
- Currency

Note: The AI can work with messy data—it will flag and handle classification issues.
</ap_transaction_data>

<supplementary_data>
If available, also provide:
- Contract database: Supplier, Start Date, End Date, Contracted Spend
- Preferred supplier list
- Category taxonomy/hierarchy
- Prior year spend for trending
</supplementary_data>

<analysis_scope>
- Time period: [e.g., Last 12 months, FY2024]
- Currencies: [e.g., All converted to USD]
- Exclusions: [e.g., Intercompany, Payroll, Rent]
</analysis_scope>
</input_data>

<spend_cube_construction>
Build the cube along these dimensions:

### Dimension 1: Category Hierarchy
Level 1 → Level 2 → Level 3
Example: Indirect → IT → Software → SaaS Subscriptions

If categories aren't provided, classify using:
- UNSPSC commodity codes logic
- Supplier name inference
- GL account mapping

### Dimension 2: Supplier Segmentation
- Strategic suppliers (>$1M or critical)
- Preferred suppliers (contracted)
- Transactional suppliers (<$50K)
- Tail spend (<$10K, >50% of suppliers)

### Dimension 3: Business Unit/Cost Center
- By department
- By location
- By project/initiative

### Dimension 4: Time
- Monthly trends
- YoY comparison
- Seasonality patterns
</spend_cube_construction>

<required_analyses>
Execute with Python and produce visualizations:

### 1. Category Deep Dive
- Treemap: Spend by L1 → L2 → L3 category
- Bar chart: Top 20 categories by spend
- Table: Category growth rates YoY

### 2. Supplier Analysis
- Pareto chart: Top suppliers = X% of spend
- Concentration risk: Spend by # of suppliers
- Tail spend analysis: # suppliers vs. spend in tail

### 3. Savings Opportunity Identification
For each opportunity type, estimate magnitude:

| Opportunity Type        | Method                                | Est. Savings |
|------------------------|---------------------------------------|--------------|
| Supplier consolidation | Reduce suppliers in category by 50%   | 5-15%        |
| Contract compliance    | Move maverick to preferred supplier   | 3-8%         |
| Demand management      | Reduce consumption 10%                | 10%          |
| Negotiation leverage   | Benchmark and renegotiate top 20      | 5-12%        |
| Payment terms          | Extend to 60 days, capture discounts  | 1-3%         |

### 4. Risk Assessment
- Supplier concentration: % spend with top 5
- Contract coverage: % spend under contract
- Single-source risk: Categories with 1 supplier
- Geographic risk: Spend by supplier country

### 5. Compliance & Governance
- PO coverage rate (if PO data available)
- Invoice exception rates
- Duplicate payment detection (if invoice # provided)
</required_analyses>

<output_deliverables>
Structure as executive presentation:

1. **Executive Dashboard** (1 page)
   - Total addressable spend
   - Top 5 categories (pie chart)
   - Top 10 suppliers (bar chart)
   - YoY trend line
   - Key metrics boxes

2. **Savings Pipeline** (1 page)
   - Waterfall chart: Current → Target with initiatives
   - Initiative table with timing

3. **Category Profiles** (per major category)
   - Spend summary
   - Supplier landscape
   - Benchmarks (if available)
   - Recommended actions

4. **Risk Heat Map** (1 page)
   - Categories by risk level
   - Action priorities

5. **Data Quality Report**
   - Classification rate
   - Data gaps identified
   - Improvement recommendations

6. **Complete Python Script**
</output_deliverables>

How to Use This Prompt

1. Export AP data: Pull 12-24 months of invoice transactions
2. Clean minimally: Don’t worry about perfect categories—AI will classify
3. Paste and run: Get instant spend cube with visuals
4. Drill down: Ask follow-ups on specific categories or suppliers
5. Build business case: Use savings estimates for initiative planning

Instant Follow-Up Analyses

• “Drill into IT spend: break down by software, hardware, and services”
• “Show me all payments to suppliers without contracts”
• “What would consolidating office supplies to 2 suppliers save?”
• “Compare this year’s marketing spend to last year by vendor”
• “Flag potential duplicate payments based on amount and date patterns”

Pro Tips

• Include GL codes even if you don’t have categories—AI can infer
• More data = better analysis (24 months enables trend detection)
• Include supplier IDs to catch name variations
• Provide your category taxonomy to ensure consistency