What it means
Screening happens at two points. Customers and suppliers are screened when they are onboarded and then re-screened whenever the lists change, while individual payments are screened as they pass through, checking payer, payee, banks and any reference details in the message.
The matching itself is fuzzy rather than exact, because names are transliterated differently, spelled inconsistently and often shared by many unrelated people. Systems therefore use similarity scoring with a threshold, and lowering that threshold catches more genuine matches while burying the team in false positives.
False positives are the dominant operational cost. It is normal for well over 95% of alerts to be discounted on review, and the resulting workload determines how many analysts a compliance team needs and how long payments sit in a queue waiting for release.
Tuning the system is therefore a permanent balancing exercise between missing a true match and drowning in noise. Firms improve the outcome by collecting better customer data such as dates of birth and registration numbers, by whitelisting names already cleared, and by testing the settings against known cases rather than adjusting them by feel.
Screening also has to be auditable. Regulators expect a firm to be able to show which lists were used, when a name was screened, what the system returned and who decided to clear an alert, so the record of the decision matters as much as the decision itself.
In practice
Real-world examples.
Example
A payment to a supplier is held because the beneficiary's name closely resembles a listed individual. An analyst compares the date of birth and country of registration, confirms it is a different party, clears the alert and releases the payment within the hour.
Example
A wealth manager re-screens its entire client book after a large addition to a sanctions list. Two existing clients now appear as genuine matches, their accounts are frozen, and a report is filed with the relevant authority the same day.
Example
A logistics firm adds vessel and port screening to its process after realising that names alone missed a restricted shipping route. The change raises its monthly alert volume by a third but closes a gap that a regulator had flagged in a prior review.
Think of it
“Sanctions screening is checking against prohibited lists-making sure you're not dealing with blocked parties.
Formula
Calculation
Two measures are commonly tracked: alert rate = alerts raised / items screened x 100, and false positive rate = alerts discounted on review / total alerts x 100.
A payments business screens 400,000 transactions in a month and the system raises 8,000 alerts, giving an alert rate of 8,000 / 400,000 x 100 = 2%. Of those, 7,960 are cleared as false positives, so the false positive rate is 7,960 / 8,000 x 100 = 99.5%.
Each alert takes an analyst about six minutes to review, so the monthly workload is 8,000 x 6 = 48,000 minutes, or 48,000 / 60 = 800 hours. At a fully loaded cost of $40 an hour that is 800 x $40 = $32,000 a month.
If careful tuning reduces alerts to 5,000 without missing any true matches, the saving is 3,000 alerts x 6 minutes = 18,000 minutes, which is 300 hours or 300 x $40 = $12,000 a month, roughly $144,000 a year.Case study
Seen in the real world.
The following is an illustrative and entirely fictional example. Northgate Payments, an invented cross-border payments provider, ran screening with a very low similarity threshold on the reasoning that catching everything was the safest approach. The result was around 22,000 alerts a month against 300,000 payments, an alert rate above 7%.
Its fictional team of six analysts could not keep up, so a backlog built and payments sat unreleased for up to four days. Customers complained, some left, and, more seriously, an internal review found that analysts under time pressure were clearing batches of alerts with a single generic note, which meant the audit trail would not have withstood scrutiny.
In this illustrative case, Northgate raised its threshold after testing the new settings against 500 historical true matches, all of which were still caught, and added structured date-of-birth data for its corporate customers. Alert volumes fell by about 60%, average release time dropped to under two hours, and the quality of the documented decisions improved because analysts finally had time to write them properly.
Watch out
Common mistakes.
- Screening only at onboarding and never again, when sanctions lists are updated constantly and an existing customer can become restricted at any time.
- Setting the matching threshold so low that the volume of false positives overwhelms the team, which paradoxically makes a genuine match more likely to be waved through.
- Clearing alerts without recording the reasoning, leaving no evidence for a regulator that the decision was properly considered.
Questions
People also ask.
Which lists should a business screen against?
At minimum the regimes that apply to its jurisdictions, currencies and banking partners, and most firms use a consolidated commercial data feed to keep them current.
Is screening only relevant to financial services?
No, exporters, shipping companies, insurers, recruiters and software providers all face the same restrictions and increasingly run screening on customers and suppliers.
What happens when a genuine match is confirmed?
The relationship or payment is frozen rather than simply refused, the authority is notified within the required deadline, and no funds are returned or released without official permission.
From the founder's library
Take it further with the book.
Build your financial confidence beyond this definition. Shihan's full-length guide, Accounting Fundamentals, takes the same plain-English approach and turns it into a complete, practical playbook for non-finance managers, business owners and students - with chapter-end quiz answers and presentation slides included.
25% off with code MMHQ25, applied at checkout. Priced in USD - checkout may show the equivalent in your local currency.
View the book and save 25%