An appropriate preliminary solution when faced with hundreds or thousands of third parties, agents, resellers and distributors is to conduct watchlist screening, which reveals data on high-risk individuals, companies and organisations. However, there is a challenge when screening against all relevant local and international watchlists: facing “false-positives”.
“False-positives” is a term to describe the possible matches that come up as part of the list of results when you run a search. For example, if you search for “John Smith”, all possible John Smiths that exist in the watchlist database will be displayed as possible matches. You can use filters and clever search strings to help you limit the search results, but it is often the case that multiple legitimate false positives come up. The question is: What is the best analytical approach to false positives?
Reasons for false positive results include:
- Full or local-language names not being known to the person searching the watchlists
- The fact that there can be many possible variations of the name you are searching
- That there can be an overwhelming number of people (or entities) with the same or similar name as your target and these are all picked up in the search.
Things to think about when conducting a false-positive analysis
1. Verifying false positives on corporate entities is relatively easy
It is not often that you end up with many unverified results for false positives on corporate entities – access to company information is relatively available, and companies within a country would be unlikely to give themselves exactly the same name. If the subject of your false positive analysis is a company, check against the company’s:
- full name and equivalents
- trade names
- license numbers and tax numbers
- date of registration
- legal form
2. Verifying false positives on individual people requires a bit more work
It is very common to find yourself with many false positives to analyse when you search on an individual. This is because it is more common for individual people to share the same name in any given country. Information about individual people is also not as readily available as those of corporate entities. If the subject is an individual, check against the person’s:
- full name and equivalents
- identification numbers (e.g. passport number or driver’s licence number)
- date of birth
- educational background.
For example, if one of the “John Smith” false positives indicated that he had been in the agriculture industry since 1978, when he was working at the United States Department of Agriculture, it is unlikely that this would be the same John Smith who is 28 years old and is a British national.
3. Right research skills
Conducting research can be a quick and easy way of verifying the false positives. It does require skills in conducting online research, however, and you may need to factor in language capabilities.
For instance, you have conducted a watchlist screening on one of your supplier company’s directors, Elizabeth Grant. The screening finds an individual named Eliza Del Grant, who is listed as a politically-exposed person as she is the former director of the Social and Family Policy Section of the Ministry of Labour and Social Affairs. The names Elizabeth Grant and Eliza Del Grant are very similar, hence further analysis is required.
Additional research uncovers a biography that was published during Eliza Del Grant’s directorship at the Ministry of Labour and Social Affairs, which says that she was born on 1 September 1950. Using your supplier company’s current corporate website, you find that it also holds a profile for Elizabeth Grant, revealing that she graduated from high school in 1968, making her approximate year of birth 1950. Further research into her Facebook profile shows her exact date of birth, which matches the one on Eliza Del Grant’s record in the watchlist.
In this instance, false-positive analysis confirmed that your supplier company’s director was a politically-exposed person. It is up to you to conduct further risk analysis to work out what to do next.
4. Ask the company or individual whether they can confirm the watchlist result
While research might be able to whittle the number of false positives down to a manageable amount (say two or three possible matches), you might simply have to approach the company or individual to see whether they can help you confirm the watchlist result.
For instance, while conducting a screening of the shareholders of a medical-distributor company in South Africa, Gamma Limited, you have a hit that shows a Behati Prinsloo who has worked at a government-controlled medical industry regulatory body located in Johannesburg.
You have found that the name Behati Prinsloo is a very common name locally. Additional research uncovers no personal details on the individual other than that the listed individual is working in a similar industry. Therefore, it is likely (but at this stage not confirmed) that the referenced individual is the subject.
In such cases, a recommended next step would be to approach the company themselves and ask them to confirm whether their Behati Prinsloo was the same as the one who has worked in the government regulatory body. People are generally comfortable with helping you confirm or deny the false positive with documentary evidence, especially if they are keen on conducting business with your company.
Keeping in mind the necessary steps of conducting false-positive analysis, performing watchlist screening remains a risk-conscious and cost-effective solution, especially for low-risk third parties, without engaging in additional on-the-ground due diligence.
In order to appropriately manage and document the processing of thousands of companies and individuals being screened, The Red Flag Group® has developed several products, including ComplianceDesktop® Compliance Technology Platform and IntegraWatch® Compliance Screening, providing a new type of data covering people and companies which have, or are suspected to have, been involved in illicit activities. In addition, The Red Flag Group® can offer support in conducting false-positive analysis.