"Where’s the Fraud?" – Combating Claim Fraud with Automation Tools and "Good Data"

To those that read my 2 previous blogs (from the 2018 Eastern Claims Conference (ECC) and the 2018 International Claims Association’s (ICA’s) Annual Education Conference), their subjects focused on how companies could use automation tools to transform their claims organizations.  Fraud was also the central theme of the 2019 Eastern Claims Conference.

Three Areas of Focus that Caught My Attention

When going through my notes of the various ECC sessions I attended (including the session that I had the pleasure of co-presenting with Rob Waterman, AVP of Fraud Operations, MassMutual), there were several key areas of focus.

While it is tough for companies to deal with the reprecussions of past business practices, future practices can and should be changed in a couple of ways:

Combating Fraud Starts from the Beginning – Identity Detection

During our session, Rob Waterman brought up how fraud detection starts right away with the new business process and knowing the identities of those involved in the policy purchase, particularly with life insurance.   The owner, insured, beneficiary(ies), payors, etc. all should have their identities verified in order to detect possible situations of fraud.  Automation tools, such as workflow, biometric detection, and integration to online identity verification services (e.g., SSDI, LexisNexis) all can come into play when verifying identities.

Knowing that the identifies were verified from the start helps remove a layer of detection when a claim occurs.  However, if past new business practices did not go to this level, then it is important to be diligent with identity verification during the claim process.

Capture Data in a Structured Manner

Throughout several conference sessions, the importance of capturing “good data” could not be stressed enough.  During new business, whatever data to be used in processes, rules, analytics, etc. needs to be captured in structured fields and the data needs to be accessible.  Also, the data needs to be reliable so that automation tools, such as workflow processes and rules, can make accurate decisions when detecting fraud. This is especially important if analytics are used, otherwise “false positives” may occur.

Unfortunately, some older policy administration systems, especially “home grown” ones present some issues on the subject of “good data.”  Having to reconcile a “good data” problem during claim time can be difficult.  Some of these issues are:

  • Capturing the same data more than once. For example, an insured has several policies, but their name is listed on one as “John Smith,” another as “Jonathan Smith,” and a third as “John A Smith.”  A lack of a centralized customer/party record is usually to blame for the situation.
  • Lack of capturing enough data. Too many times I hear of how not enough beneficiary information was originally captured long ago and how that presents a problem during claim time.  Identity verification is difficult or even tracking down the beneficiary can be impossible.  This makes fraud detection very complicated.
  • Lack of awareness of other lines of business or types of insurance. Some companies have many, many administrations where particular niche product is administrated on a particular system. A variable universal life (VUL) product administrated on Policy System A may not be “aware” of other products on other systems come to claim time.  The lack of sharing of information may not accurately trigger financial based fraud rules where an entire incident’s claim value is looked at.

Bringing it all together – the Centralized Claim Case Management System

A modern claim case management system can help solve many, but not all the pitfalls, with fraud detection especially when it is tightly integrated with workflow, rules, and analytics.  It is especially important for the claim system to be communicate with many other sources of data:  one or more policy administration systems, the online “datamarts,” internal data warehouses, etc.  An organization that brings it all together can benefit in many ways:

  • Document artifacts – being able to centralize claim documents for the “event” (e.g., death, disability, etc.) helps the user record structured data once in a consistent manner, use the documents to adjudicate possible several policies that may respond to the event versus adding said documents over and over, and establishes a “single source” whether the event is valid or if any fraud is involved.
  • Single source of truth – did you ever hear the adage “singing from the same hymn book?” A centralized claim system establishes this so that there is one place to respond to a claim event or to access past claim events for an entity.  I have been asked many times over the years “Can I see past claims for the insured, such as a waiver or critical illness?” or “Can I see all claims that we’ve paid a medical provider?”  With a centralized claim system, the answer is “yes.”  Also, workflow, rules, and analytics with good, reliable data points could be configured to detect automatically if anything triggers a possible fraudulent event.

The right tool and good data make all the difference

In conclusion, we need to remember that automation tools are merely that:  tools.  Have you ever tried to use the flat-head screwdriver when you really needed a Phillips screwdriver?  Or, tried to saw something where the saw’s teeth aren’t very sharp?  The same goes for automation tools and claims software.  The user can only do the job so well based upon the tools they are given.  However, with claims automation tools, an automated fraud detection process, its rules, the analytics, etc. is only as good as those that configure and “program” the tools.  And, at its core, it starts with good, reliable structured data that where specific data points are detected as the known basis of possible fraud patterns.  When companies are able get reliable data and have it in an accessible, structured format, they can then be off to a great start in the world of automated fraud detection.

To read more about FINEOS Claims, please visit our website.

You may also be interested in