Analytics: Inside the New Fraud-Fighting Mindset

April 24, 2026 / 7 min read

Young insurance investigators and law enforcement officers often want to be in the field, inspecting losses, examining evidence, conducting face-to-face interviews, and catching the bad guy. I was once that law enforcement guy: at the ready, wanting to chase the villains instead of sitting behind a desk looking at spreadsheets and running data queries in different software applications to identify patterns. Somewhere along the way, my mindset changed. Strong parallels emerge when comparing the power of analytics in police work and insurance investigations. First, consider the story of the bank robbers.

analyst

Robbery in progress

A young officer on patrol responded to a call of an armed robbery in progress. Most robberies happen very quickly, and the criminals are gone before the police reach the scene, unless it’s on your favorite television show. In this instance, the officer was very close, arriving just as the suspect was fleeing.

A vehicle pursuit and a foot chase ensued, and the armed suspect was caught and arrested. The money and the robber’s ski mask were recovered on the scene. There were no serious injuries, and the law enforcement officers celebrated the removal of one more violent offender from the streets. The robber was easily convicted at trial.

Two years later, that same officer was assigned to an undercover fugitive team working to locate and arrest the most violent criminals. On a list of recent parole violators, he recognized the armed robbery convict he had arrested just two years ago. Convicted of multiple felonies, including armed robbery, fleeing and eluding, and possession of a firearm, the individual was already released and even had time to violate the terms of his parole in less than 24 months. This cycle of recidivism is disturbing but common.

The power of analytics

Over many years, that same officer learned the power of analytics, starting with cellphone geolocation and call chaining. Following that was Network Analytics—connecting individuals by phone numbers, addresses, emails, and social media contacts. These skills were integral in making sense of large quantities of data and extracting the story behind the numbers.

The officer began to work a string of armed robberies in his area with no initial suspects. One anonymous license plate tip pointed to a vehicle owner who was a paroled felon. Working with the parole officer led to a cell phone number and then phone records encompassing the times when the armed robberies occurred.

The phone records showed the suspect’s phone was in the area of eight out of nine of the armed robbery locations. Analyzing other numbers the suspect spoke with just before the robberies led to two more suspects, whose phones showed similar activity. All would converge near their target about 30 minutes before the robbery, immediately leaving the area afterward. Link analysis connected all three subjects through prior arrests and social media contacts, in addition to their call records. The officer assembled his findings into an intelligence package for federal prosecutors.

All three subjects were brought in for questioning. Due to the intelligence gathered, all ended up confessing to twice the number of armed robberies for which they were originally suspected. All were convicted and sentenced to a minimum of 30 years in federal prison with no chance of early parole. There was no vehicle pursuit, no foot chase, and no shots fired. The public wasn’t endangered, and no officers were put in harm’s way to repeatedly arrest the same violent offenders. The effect of this analytical mindset is far-reaching. It can be the same for insurance investigations.

Hijacked policy, low-touch claim

An adjuster with Insurance Company A receives a claim. The insured states he was driving his vehicle when he swerved to avoid a dog—and instead, hit an individual riding an e-bike. A long-time customer of the carrier, the insured is apologetic and has almost no prior claims history. The adjuster sees it as abnormal that the insured changed both his email and phone number when reporting the claim. However, this adjuster is juggling many claims, and the loss is less than $5,000, with no injuries. It’s decided to move to resolve the claim quickly and make an electronic payment. The claimant with the wrecked e-bike provides photos, an email address, and a phone number, and the claim is promptly paid and closed.

Hijacked policy, analytic mindset

The same scenario happens at Insurance Company B, with the same details. The adjuster there also sees it as strange that the insured changed his phone number and email at first notice of loss (FNOL). This adjuster and his company have an analytical mindset and decide to dig deeper before paying the low-exposure claim.

This adjuster calls the original number (not the FNOL number given) and reaches the insured, asking for details about the e-bike claim. The insured says he has no idea what the adjuster is talking about, and his vehicle hasn’t been in an accident. The adjuster then calls the phone number given at FNOL. It’s clear the voice that answers and alleges to be the insured has a significantly different pitch and speech pattern from the voice at the original number documented in the policy.

The adjuster contacts his SIU Link Analysis using Verisk’s solution, formerly known as NetMap and driven by the ClaimSearch^® database, finds the phone number given at FNOL is connected to dozens of similar claims with different names in the past six months. Further, the SIU requests photos to validate the claim. Reverse image searches of the damaged e-bike photos show they were taken off the internet. As a subscriber to Digital Media Forensics, the SIU also learns those photos had been used on recent similar claims at different insurers in different states.

Close to 50 claims are identified as being linked to this organized claims fraud. Synthetic and stolen identities were used, and Company B adds those identifiers to its internal advisory list in the event its adjusters ever encounter those phone numbers, email addresses, names, or other information again. This SIU will be alerted and protected from similar fraud in the future.

Analytics develops the big picture

Analytics can alert investigators and analysts to non-obvious relationships and connections that shouldn’t exist. For example, when investigating a traffic crash claim, it can be quickly realized that the driver of the insured vehicle in today’s claim has been involved in a prior traffic crash claim where they were an occupant in the claimant’s vehicle. If the insured and claimant deny knowing each other when asked, their denial provides evidence of material misrepresentation.

Analytics support the mission

Insurers want to pay all that they owe. They don’t want and shouldn’t have to pay what they don’t owe. Paying fraudulent claims in a rush to settle them does a grave disservice to consumers who trust their insurer. Those insurers have a responsibility to ensure their customers get the most value possible from their premium dollars.

Analytical tools are available to help quickly sort valid claims from suspicious ones. Failing to use these tools and adopt the analytical mindset can be put carriers at a competitive disadvantage and contribute to the insurance fraud problem in the U.S.. Analytics may not present the easiest path, but these tools can deliver the best and most accurate results. The effort is worthwhile, and the costs are quickly recovered.

Tracking the amount of invalid claims identified and ultimately not paid shows how analytic tools justify their cost. Spending premium dollars on fraudulent claim detection is far more ethical and desirable than paying premium dollars on fraudulent claims in an effort to resolve them quickly.

Evolution of fraud detection techniques

As fraud evolves, so must the techniques to identify it. Hijacked policies combined with synthetic or stolen identities call for a new mindset and more advanced techniques to address this newest evolution of fraud. It’s a delicate dance between paying claims quickly and stopping questionable scenarios for a second look. If insurers use this technology effectively, they can become far more efficient at identifying irregular claim patterns and empower themselves to be more responsible stewards of their customers’ premium dollars.

Learn more about how Verisk’s comprehensive solution suite is designed to help detect fraud faster.

Learn more