Methodology

Nigeria Procurement Watch analyses publicly available federal procurement data from the Bureau of Public Procurement's NOCOPO portal. This page explains exactly how findings are generated, what they mean, and — just as importantly — what they don't mean.

What counts as a contract

NOCOPO records procurement at four stages: Planning, Tender, Contract, and Implementation. We only analyse contracts in the Contract or Implementation stage — meaning an award has actually been made. Planning-stage budget allocations (which are ceilings, not awards) are explicitly excluded from all scoring and detection.

Detectors

Price Inflation

Flags contracts priced far above the median for that agency, with extra scrutiny removed for capital purchases (vehicles, buildings) where large one-off costs are structurally normal.

Single-Source Concentration

Flags vendors receiving a disproportionate share of one agency's total contract value.

Split Contract

Flags multiple near-identical-value contracts to the same vendor, which can indicate deliberate splitting to avoid procurement thresholds.

Duplicate / Decimal-Shift Pattern

Flags pairs of similar contracts where one value is ~100x the other — which may indicate a data entry error or a genuine duplicate submission. We present both possibilities.

Vendor Verification

Flags high-value contracts to vendors with no other record in our current dataset. This is a prompt for manual CAC verification, not an accusation — our dataset is not yet complete, so single appearances are expected for many legitimate vendors.

Scoring

Each agency's Anomaly Risk Score (0–100) is calculated from the severity-weighted density of findings relative to its total contract volume — not a raw count. This prevents agencies with more contracts from being penalised simply for having more data. Agencies with fewer than 3 awarded contracts are capped at a low score regardless of findings, since there isn't enough data to draw a confident conclusion in either direction.

What this is not

This dashboard does not allege corruption. It surfaces statistical patterns in public data that warrant further investigation by journalists, civil society, or the agencies themselves. Many flagged patterns have innocent explanations — a data entry error, a legitimate capital purchase, or incomplete data coverage on our end. We try to make that uncertainty explicit in every finding rather than implying false certainty.

Data source

All contract data is sourced directly from nocopo.bpp.gov.ng, the Bureau of Public Procurement's public Open Contracting Data Standard portal.