Back to Blog

In the world of identity verification and fraud prevention, seemingly small details can make an enormous difference. At Socure, we’ve discovered that address verification is one such detail that requires far more sophistication than most realize.

The Challenge: When “Sources of Truth” Fall Short

As part of the Customer Identification Program (CIP) process, financial institutions must not only verify a customer’s identity but also form a reasonable belief that they could locate the individual in the event of an Anti-Money Laundering (AML) investigation. This is where accurate address verification becomes critical.

PO boxes and Commercial Mail Receiving Agencies (CMRAs) present a particular challenge for this requirement. When applicants provide these addresses instead of physical residential locations, it undermines the institution’s ability to reasonably locate them if needed.

Logically, the United States Postal Service (USPS) should be the definitive source of truth for identifying post office boxes and postal facilities. After all, they create and maintain these addresses. However, our extensive testing revealed a startling reality: even the USPS’s own API services have significant blind spots in their data.

Case Study: The Invisible Post Office

Consider this real example we encountered: an address at “9766 Highway 550” in Counselor, New Mexico. On the surface, this appears to be a standard street address. However, a Google search revealed this location is actually a post office.

When we attempted to verify this through official USPS channels, we hit a wall. 

Despite this location appearing on the USPS website, their API services didn’t identify it as a post office. Even more surprisingly, searching for postal facilities within the same zip code (87018) failed to return this location – whether we searched within a 5-mile radius or expanded to 50 miles.

How could a postal facility be visible on the USPS website but invisible through their own search and API services?

Socure’s Solution: Going Beyond Standard APIs

This discovery led us to develop a more comprehensive approach:

  • Multiple data sources: We recognized that even authoritative sources have gaps, so we implemented cross-referencing across multiple data channels.
  • Algorithmic iteration: Rather than simple API calls, we created systems that iterate through all possible zip codes and location IDs to find USPS facilities and post office locations that might otherwise be missed.
  • Cross-verification: We applied similar techniques to other shipping and commercial mail services, including FedEx, UPS, and DHL.
  • Intelligent reclassification: Separately, we identified that many addresses were being incorrectly flagged as commercial simply because applicants omitted unit numbers in mixed-use buildings. By checking against our identity graph, we could see if the same address with a unit number was classified as residential and reclassify accordingly.

The Results: From 0.5% False Negatives to 0%

Our enhanced methodology dramatically improved accuracy. For commercial mail receiving agencies and PO Box flagging alone, we reduced the false positive rate from 0.5% to 0%. This means flagging the riskiest customers in the right context.

The Bigger Picture: An Authoritative Framework

This experience underscored a key principle at Socure: different sources vary in reliability for specific PII elements. For example, the Social Security Administration is authoritative for SSNs but less accurate for date of birth or first name matching due to fuzzy matching tolerances. USPS may classify an address as commercial if the unit number is missing, and the VA may not find a record if an applicant hasn’t updated their address.

By understanding these nuances and building systems that account for data deficiencies, we’ve created verification processes that are both more accurate and more user-friendly.

  1. Even official “sources of truth” have significant data gaps
  2. Effective identity verification requires sophisticated approaches beyond simple API calls
  3. Cross-referencing multiple sources creates more reliable outcomes
  4. Understanding the specific strengths and weaknesses of each data source is essential
  5. Small improvements in data accuracy can dramatically reduce friction for legitimate users

At Socure, we’re committed to continuous innovation in our data methodologies, ensuring our clients can confidently meet compliance requirements while delivering seamless experiences to their customers.

Josh Linn
Posted by

Josh Linn

Topics:
Josh Linn

Leading digital identity verification and authentication strategy for a top 10 FI before joining Socure in 2018 to lead Data Acquisition efforts, Josh Linn now serves as Senior Vice President of Machine Learning Product Management & GM of RegTech leading innovation for the regulatory compliance and predictive analytics platforms. He holds and MBA from Syracuse University and a Master's of Science in Information Systems from Northwestern University.