New

Nigeria import & export data now updated weekly. See coverage →

Methodology

No black box. Every record traces to a source.

Trade intelligence is only useful if you can defend it in front of a credit committee, a regulator or a board. This page documents exactly how our data is obtained, processed and published — including where it falls short.

100%
Records with source tag
Weekly
Publication cycle
48 hrs
Typical filing-to-publish lag
HS 2022
Classification base
Pipeline

Four stages, each independently auditable.

Every record moves through the same four stages regardless of market, and carries a processing history you can inspect.

01
Collect

Official customs releases, bills of lading and contracted partner feeds are ingested in their native format with the original document retained.

02
Clean

Encoding, transliteration and formatting are normalised. Obvious filing errors are flagged rather than silently corrected.

03
Enrich

Entities are resolved, HS codes harmonised, ports mapped to UN/LOCODE, and corridor context attached.

04
Publish

Records are indexed for search, versioned, and released with a source date and confidence indicator.

Controls

The quality controls that sit between filing and search result.

Provenance tagging

Each field carries the source it came from, so a value derived during enrichment is never confused with a filed value.

Outlier detection

Weights, quantities and container counts are screened against heading-level distributions; extreme values are flagged, not deleted.

Entity resolution

Company matching uses name, address and behavioural signals with a confidence threshold. Below threshold, records stay separate.

Classification review

Ambiguous descriptions are mapped conservatively and marked, so heading-level totals are never inflated by guesswork.

Restatement policy

When an authority revises a filing, we republish the record and retain the prior version rather than overwriting history.

Sampling audits

A rolling sample of published records is manually reconciled against source documents each cycle.

Sourcing

Where African trade data actually comes from.

African trade records are fragmented across customs authorities, port community systems, shipping manifests and partner networks. Each market publishes a different subset with different granularity and cadence.

  • National customs authority releases and gazetted trade statistics
  • Bill-of-lading manifests obtained through licensed channels
  • Port community system records where publicly available
  • Company registry data used only for entity enrichment
  • Contracted commercial partners for markets without open publication
Customs officers and clearing agents reviewing shipping documentation
Limitations

What this data cannot tell you.

Coverage is not universal

No commercial dataset captures every African consignment. Informal cross-border trade, certain free-zone movements and some bilateral arrangements are outside the filing systems we draw from. Where a market's coverage is partial, the country page says so explicitly.

Declared values are not market prices

Where values are published they reflect what was declared for customs purposes. They are useful for relative comparison and anomaly detection, but they are not a price index and should not be used as one.

Classification reflects the filer

HS codes are assigned by the party lodging the declaration. We harmonise and flag, but we do not reclassify goods on their behalf — doing so would replace a documented decision with our own assumption.

Lag varies by market

Publication lag ranges from around 48 hours to several weeks depending on the authority. Current lag per market is published continuously on the platform status page.

FAQ

Questions, answered.

Is the data legally obtained?

Yes. All records are obtained from official publications or through licensed and contracted channels. We do not scrape restricted systems.

Do you sell personal data?

No. Our records concern commercial entities and consignments. Where a filing contains personal information, it is removed before publication.

Can I get the raw source document?

Enterprise customers can request source-level verification for specific records as part of a due-diligence workflow.

How do you handle corrections?

Report a suspected error and we will re-check it against source within one publication cycle, then republish with a versioned correction note.

Data you can put your name against.

Talk to our data team about coverage, sourcing and verification for your specific markets.