Commodity Phase-1 Brief
Phase 1 narrows the project to FAO / FAOSTAT only. The goal is to build the cleanest possible FAO-based extraction and analysis plan for coffee, tea, cocoa, and—if the item coverage is strong enough—mate.
1. Phase-1 source decision
Primary source: FAO / FAOSTAT
This phase focuses on what can be extracted, modeled, and analyzed from FAO alone before adding other sources later.
- Use cocoa, not chocolate, as the agricultural commodity concept.
- Treat mate as a real candidate if FAO item coverage is strong enough.
- Deprioritize matcha for Phase 1 because it is more naturally handled as a tea subcategory than a standalone FAO commodity.
2. Comparison dimensions
Production
Country-year production quantity and unit for coffee, tea, cocoa, and possibly mate.
Harvested area + yield
Area and yield make the comparison more explanatory and help distinguish output growth from productivity growth.
Trade
Use FAOSTAT trade tables only if the coverage and field cleanliness are strong enough for the selected commodities.
Phase-1 note: price is no longer the lead layer if the project is FAO-only. Price can return in a later phase with a different source if needed.
3. FAO-only source/usefulness assessment
| FAO area | Usefulness | Best role | Notes |
|---|---|---|---|
| FAOSTAT Production (QCL) | High | Primary production backbone | Best source for country-year commodity production, harvested area, and yield. |
| FAOSTAT Trade (TCL) | Medium-High | Trade layer if coverage is clean enough | Useful, but should be validated commodity by commodity before making it core. |
| FAO definitions / metadata | High | Classification guidance | Critical for confirming item meaning and avoiding concept drift. |
4. What Phase 1 should answer
- How do coffee, tea, cocoa, and possibly mate compare in production over time?
- Which countries dominate each commodity?
- How do harvested area and yield differ by commodity and country?
- If trade coverage is strong enough, how do export/import patterns differ across the commodities?
5. Recommended Phase-1 data model
production_country_yeararea_yield_country_yeartrade_country_year(only if FAO trade quality is sufficient)commodity_lookup
Preferred architecture: country-year first, with separate tables instead of one oversized combined table.
6. Scope options
Best Phase-1 scope if we want the least ambiguity and the strongest comparability.
Strong candidate if mate is confirmed usable enough in FAO item/domain coverage.
7. Next hard step
Build an exact FAOSTAT item/domain map for:
- coffee
- tea
- cocoa
- mate
Then confirm which measures are actually available and clean enough to drive Tableau dashboards.