Analaysis of risk in a claimant's voice can start the process of eliminating genuine claims and move the process on to identifying the fraudster. Kerry Furber explains

To download the Power Point presentation for this article please
click here (549k)

Contrary to the beliefs of some, Digilog's AVS (advanced validation solution) process is a multi-layered, systematic process, combining what we think is the world's most powerful voice risk analysis (VRA) technology, plus advanced questioning skills and narrative integrity analysis.

The process focuses on the actual claim itself, as reported by the customer in real time. It does not use more commonly deployed screening techniques that focus on such things as a customer's socio-economic circumstances, demographics and prior insurance history.

When the VRA technology is used, a claimant's voice stress levels are calibrated according to questions that can only be answered honestly by the customer; hence the pre-existing stress level of each individual customer is taken into account. From this baseline, specific variations (that may indicate a notable risk) can be identified.

Digilog's AVS also uses highly advanced, manual techniques that can identify significant risk in conversational responses.

Corroboration between the two methods is then sought, thus determining whether a claim is low risk (probably genuine). Such claims are then moved swiftly on to a fast track for settlement (the primary objective), thus eliminating the often unwarranted, customer-frustrating and costly realm of the claims investigation.

In the past 12 months, about 72% of claimants screened via Digilog's process have been moved onto the fast track within just a few minutes of registering their claim.

The 'pure' process has three main stages. In stage one, call centre staff swiftly identify genuine claimants and pass these cases for fast track closure of the claim. They are assisted by the following:

  • VRA technology - used in specific areas of a scripted conversation
  • Narrative integrity analysis (NIA) - a process aiding the identification of truth/deception indicators in conversations
  • Conversation management skills
  • Applied psychology, based on the principle that genuine customers have no difficulty recalling events or details about something that really happened, while, in contrast, the fraudsters struggle.
  • Claims that indicate high risk are then retained for stage two screening and further validation. In stage two, trained claims handlers conduct an in-depth but customer-focused probing of the claim.

    Specific risk issues will have been flagged up at the primary stage, enabling focused exploration with these key objectives in mind:

  • Reassessment of the original risk decision and return any low risk claims back to fast-track for settlement
  • Checking for anomalies between stages one and two information capture
  • Unnerving the fraudster, by encouraging him to either drop or change his claim (commonly and painlessly over 50% actually do this)
  • Obtaining information that enables defensible rejection of fraudulent claims
  • Production of focused reports on risks indicated and propose cost-effective lines of investigation that may secure a repudiation via tangible evidence.
  • In stage three all remaining suspicious cases are investigated according to the client's practices. At the conclusion of this stage, individual case handlers will be able to review all accumulated evidence and make defensible decisions to repudiate or settle according to the burden of proof in civil law.

    It is important to emphasise that under no circumstances are decisions to pay or deny a claim taken solely on the basis of VRA technology. All decisions are based on tangible, traditional evidence and are taken following a focused investigation - now enabled by the Digilog AVS's ability to flag up very specific, high-risk areas.

    This must be capable of withstanding the scrutiny of a court of law. Hence it is virtually impossible to be a genuine claimant yet have a claim denied on the grounds of fraud.

    Trials in the past year have seen one insurer improve their fraud-related defeat rate from around 1% to 19%. Additionally, claims handling times from notification to settlement are plummeting - by around 75% in general.

    Correspondingly, the effect on dissatisfied customers has been equally as dramatic, reducing customer complaints to well below 1%. Digilog regards this as true testimony to the success of its AVS process and re-iterates that genuine customers have nothing to fear and much to gain from this contemporary and highly efficient solution.

  • Kerry Furber is managing director of Digilog
  • Topics