Well, the most common issues with Sanctions and PEP scanning are: costs and the volume of false positives. What can you do about it?

First, let’s talk about the cost. 💸

There are two general ways of paying for access to sanctions scanning and usually the choice depends on the number of customers you have and how many development resources you are ready to allocate to developing in-house compliance solutions:

• When your startup is new, resources are scarce and timing is critical. Many startups prefer to send their list with customer data to a scanning provider, and get the response back, which is either a “clear” or “XX% match”. This solution is more suitable for younger startups because the implementation is faster and because the number of customers they have is not sufficient to justify more complex in-house development efforts.
• More mature startups prefer to buy access to the sanctions lists and build their scanning and matching logic in-house. It takes longer and requires specialized know-how, but it’s more scalable.

How does it impact pricing?

Pricing becomes critical when you need to do periodic scanning of all your customers. Just imagine, if you have to scan 200k customers every month and it costs you 30-50 cents, it gets pricy very quickly.

• When you send the names of your customers to the data vendor who provides scanning services for you, your pricing will likely be linked to the number of customers, which makes it less scalable.
• When you purchase access to the sanctions lists directly and you pay for access to these lists and for all periodic updates; additionally, you have to invest into building scanning and matching logic in-house which makes your pricing agnostic to the number of customers and therefore more scalable.

What about false positives?

Handling false positives is the big and ever-growing elephant in the sanctions and PEP scanning room. Why? Because if you don’t handle it well (by that I mean automation), this process will require more and more manual resources and will generate more and more audit issues and backlog tasks.

Which is why if your startup is reaching 100k registered customers, I strongly recommend you invest time and efforts into a scalable and automated scanning solution with the following features:

• Basics: Setting up reliable auto-dismissal logic that can handle partial matches, reconciliation of the countries of residence, dates of birth (or missing dates of birth), transliteration intricacies with Chinese and Arabic names, and all sorts of issues with common names – this is not trivial and requires specific know-how.
• Advanced: You need to train your alerts and hits generating tools to learn from the results of past manual reviews and continuously improve your scanning logic, otherwise you will be swamped by repetitive and redundant hits generated by various tools and systems. For example, if you have a number of linked accounts under common control or several accounts sending funds to the same beneficiary (it could be a merchant or a charity, for example), your systems must be trained to remember and recognize that this alert has already been manually reviewed and dismissed in the past, and therefore redundant repetitive alerts should be eliminated whenever possible.

Adverse Media Scanning 

It should be noted that technically speaking adverse media scanning is not a regulatory requirement. It might be a great source of information for keeping your customer details up-to-date, however, you really need to know what you are doing.

If you are a young startup, I recommend you to be very selective and potentially steer away from integrating this feature, unless you really know why you need it. Most articles or social media scanning solutions don’t come with perfect filtering, which is why this feature can generate a lot of false positives and “empty” alerts, which may require disproportionate efforts from your team to manually review and dismiss them.

What has been your experience with scanning solutions thus far? What works best for you and what doesn’t?

Happy to hear your thoughts on this! 💭