When it comes to fighting document fraud, lenders should have “Great Expectations.”
We are not talking about Charles Dickens’ 19th century novel. Rather, we are referring to the ability to quickly distinguish between legitimate and illegitimate income documents using a concept we call “Expectation Definitions.”
What is an Expectation Definition?
An “Expectation definition” means the insight one gains into a data domain after one has seen sufficient examples. A professional baseball hitter can quickly recognize a pitch upon release only after seeing thousands of change-ups, fastballs, sliders, and curves. Elite tennis players can return a 130mph serve because they have built up their muscle memory over decades to respond instantaneously. An experienced plumber or repairman or doctor can diagnose a problem only after a few questions. In each case, domain expertise from examples leads to near-instant recognition.
Our Expertise
At Informed, we have built a robust expectation definition for bank statements, pay stubs and other income documents after reviewing tens of millions of income documents. Our system has processed documents from:
- Over 1900 banks
- More than 1.2 million employers
- Over 85,000 auto dealers.
How Expectation Definitions Detect Fraud:
So, how does “expectation definition” provide value in mitigating document fraud?
- Identify variations from standard templates: We know the standard template for thousands of banks and employers. For each template, we know the minute details about the document including the font used, the version of the document, the positioning and shaping of logos and other visual assets, the formats of dates, where signatures are placed, and the arrangement of information on the document. So, when a fraudster submits a document which varies from a lender template on any dimensions described above, the variation is detectable. For example, a fraudster could put date formats xx/xx/xxxx versus xx-xx-xxxx, or fail to put a $ on a paystub or simply capitalize the first letter of an individual’s name when the template uses ALL CAPS.
- Identify lender specific variations: Lenders often use more than one template, and we can see the variations in the documents they use. For example, many lenders have different templates for individual accounts, joint accounts, business accounts, English accounts, Spanish accounts, paper statements, and electronic statements. Fraudsters who use the wrong format will reveal themselves.
- Quickly identifying new document versions: If a lender changes its template or creates a document, we notice it nearly instantly. Because of the constant rush of live documents through our system, we’ll see multiple examples of new documents within hours and thus can confirm its validity. We also have an expectation, based on prior experience, how often versions of the document change.
A robust expectation definition minimizes fraud and can improve the customer experience. It reduces fraud because it enables companies to pinpoint anomalous documents quickly. It also helps lenders reduce friction by allowing legitimate documents to be rapidly recognized. This quick recognition allows lenders to bypass second level reviews so the document can speedily move through the approval process.
Lenders should have great expectations of their fraud solutions. And this includes great expectation definitions.
