A recurring conversation: a client has run their standard integrity screening on a West African counterparty, the report came back clean, and something has since gone badly wrong. The question is how the process missed it.
In our experience the answer is almost never that the screening was done carelessly. It is that a methodology built for jurisdictions with deep, digitised, well-indexed public records was applied unchanged to one that does not have them, and the absence of findings was read as a finding.
One: database coverage mistaken for market coverage
Global screening providers are aggregators. Their coverage of any given jurisdiction reflects what they have been able to license, scrape or partner into. For much of West Africa that coverage is thin, particularly for litigation, insolvency and regulatory action.
The practical consequence is that "no adverse findings" and "our sources do not extend to this jurisdiction" produce an identical-looking output. The remedy is to ask the provider directly what their source coverage is for the specific country, and to treat the answer as part of the report.
Two: registry data assumed to be current
Filing obligations exist across the region and are inconsistently enforced. It is entirely normal to find a commercial register entry showing a shareholding and board composition that is several years out of date, because nobody compelled an update and there was no commercial reason to volunteer one.
A register extract is evidence of what was filed, not evidence of the current position. Corroborate against something else: recent contracts, bank mandates, audited accounts, or a direct representation from the counterparty that you can hold them to.
Three: the language gap
Adverse media screening in English will systematically miss Francophone West Africa. The relevant reporting is in French, much of it in regional outlets that are poorly indexed by international aggregators, and some of the most useful material is in local-language or community media that is not indexed at all.
Court and gazette material is also in French, and machine translation of legal French is reliable enough to triage but not reliable enough to conclude on. If nobody on the team reads the source language, the search is not being done.
Four: ownership tracing that stops at the first layer
Identifying the registered shareholder of the local operating company is the beginning of beneficial ownership work, not the end of it. Where the first layer is a holding company in another jurisdiction, the trail frequently continues through structures specifically chosen for opacity.
Two habits help. First, keep going until you reach a natural person or until you can state precisely where and why the trail stopped. Second, treat the choice of intermediate jurisdiction as information in itself — a structure with no operational or tax logic is usually there for a reason worth understanding.
Five: no human corroboration
The single highest-yield step in frontier-market diligence is also the one most often skipped: speaking to three or four people who have actually transacted with the counterparty. Former distributors, competing suppliers, sector lenders, local counsel who has been on the other side.
This surfaces the categories of problem that never generate a document — payment behaviour, how disputes get handled, whether commitments survive a change in circumstances, who really makes decisions. Done properly it is lawful, ordinary professional enquiry. Done improperly it is pretexting, which is neither, and the distinction is worth being strict about.
What a good report looks like
Three characteristics, none of them expensive:
- Sourcing is visible. Every material claim carries where it came from, so the reader can weigh a registry document differently from a market conversation.
- Confidence is stated. Distinguish established fact from reasonable inference from unverified allegation. These are three different things and a report that blends them is not usable for a decision.
- Gaps are declared. "We were unable to establish beneficial ownership beyond the intermediate holding company, because that register is not public" is a more valuable sentence than silence in the same place.
The purpose of diligence is not to produce a clean report. It is to tell the decision-maker what is known, what is not, and how much weight each part will bear.