NaijaMarket Intel was built to solve a specific problem: Nigerian food prices are opaque, fragmented, and often misleading. A trader in Kano and a buyer in Lagos can be working from prices that are days old and sourced from a single reporter. We built a three-layer verification architecture to make our data verifiably accurate — not just plausible.
Verification Pipeline
GPS-Verified Field Submission
Approved traders submit prices directly from physical market stalls via WhatsApp. Each submission is geo-fenced: the trader's GPS coordinates must fall within 500 m of the declared market. Submissions outside this radius are automatically rejected before reaching the database. Each trader carries a reputation score (0–100) updated with every accepted or rejected submission — traders scoring ≥ 80 receive instant approval.
3-Validator Consensus
Every submitted price is independently reviewed by three validators assigned to that market. Validators compare the submission against our rolling baseline (EWMA of verified historical prices) and vote approve or reject. A submission requires majority consensus (2 of 3) to be marked APPROVED. Price variance > ±30% from baseline triggers an automatic fraud flag and escalates to senior review regardless of validator votes.
Confidence Scoring
Each price in our database carries a confidence_score (0–100) computed from four factors: consensus margin (unanimous vs split vote), validator accuracy ratings, submitting trader's reputation score, and distance from the item's 30-day baseline. A score of 85+ is considered high-confidence. Consumers with BUSINESS tier and above see this score alongside every price so they can weight their decisions accordingly.
3× Daily Generation
Verified submissions feed a three-slot daily price generation pipeline running at 08:30, 11:30, and 14:30 WAT (West Africa Time). Each slot generates 137,860 price points across all item-market combinations. Markets with verified submissions for a given slot use REAL_ANCHORED pricing. Markets without a fresh submission carry forward a simulated estimate (SIM_TRACKED) anchored to the last verified price, clearly flagged in the data. A watchdog function detects and backfills any missed slots automatically.
NBS Cross-Reference & NFPI
Our platform maintains a parallel reference dataset sourced from the National Bureau of Statistics (NBS). NBS items are excluded from consumer-facing prices but used to calibrate the NaijaFood Price Index (NFPI). The NFPI is a Laspeyres-type index with base period January 2016 = 100, computed monthly across a fixed basket of 44 food commodities weighted by Nigerian household expenditure patterns. The current NFPI reflects the April 2026 CPI rebase (NBS inflation rate: 16.06%).
Data Freshness
Prices are generated three times daily. Each generation slot produces 137,860 price points (610 items × 226 markets). The price_date field on every record reflects the date of generation. Consumer-facing APIs always serve the most recent slot. Our Latest_Prices_Summary view is refreshed before each morning slot and after each generation run.
Coverage
We track 610 commodity items across 226 markets in 37 states and the FCT. Coverage spans the six geopolitical zones with market density proportional to trading volume. Below is the breakdown by category:
| Category | Sample Items | Markets |
|---|---|---|
| Grains & Cereals | Rice, Maize, Millet, Sorghum, Wheat flour | 226 |
| Tubers & Roots | Yam, Cassava, Sweet potato, Cocoyam | 267 |
| Vegetables & Tomatoes | Tomatoes, Onions, Pepper, Spinach | 254 |
| Proteins | Beans, Fish, Chicken, Beef, Eggs | 231 |
| Oils & Fats | Palm oil, Groundnut oil, Soya oil | 248 |
| Cash Crops | Groundnut, Sesame, Soybean, Cashew | 176 |
Data Availability by Subscription Tier
- Live prices
- 7-day history
- All markets
- + NFPI access
- Basket tool
- Price alerts
- + Price forecast
- Market snapshot
- Full history
- + Cross-state arbitrage
- CSV/Excel export
- Full analytics
- + Multi-user access
- API access
- Confidence scores
- + Custom SLA
- White-label API
- Raw data feed
Historical Archive
We maintain a 10-year price archive (2016–2025) in Azure Blob Storage as compressed Parquet files, accessible via the ENTERPRISE data feed. The 24-month hot window (current month minus 24) is queryable in real time via API. Historical data uses the same item and market IDs as live data, making longitudinal analysis straightforward.
Known Limitations
Prices are indicative, not transactional. NaijaMarket Intel reports observed market prices from physical traders. Prices on our platform reflect the conditions at the time and location of submission. Actual transaction prices may vary based on quantity, buyer relationship, and negotiation. Our data should not be used as the sole basis for procurement or trading decisions.
Markets with no recent verified submission use SIM_TRACKED estimates, clearly flagged in the data_source field. Estimates are anchored to the last verified price adjusted for recent trend signals.
Data Queries & Enterprise Access
For data licensing, API access, custom SLAs, or questions about our methodology, contact our data team:
- Data inquiries: data@naijamarketintel.ng
- Enterprise sales: enterprise@naijamarketintel.ng
- General support: support@naijamarketintel.ng
- WhatsApp: Chat with us on WhatsApp