Business challenge

Fraud in the insurance industry raises costs for everyone. How can leading insurers move faster than the criminals to block fraudulent claims before they are processed?


This insurer uses automated alerts to send predictive indicators of fraud to investigators in real time—enabling them to build strong cases against fraudsters before they have a chance to claim.



early warnings that give fraud investigators a head start


investigators to build stronger cases against criminals, faster


asset recovery costs, delivering significant savings

Business challenge story

Beating the clock to prevent fraud

Fraud is one of the biggest cost drivers in the insurance industry. To keep premiums competitive for consumers, reducing the risk of fraudulent claims is moving to the top of the agenda at many leading firms.

Combating fraud is a key goal for this leading insurer. The estimated cost of fraud for the US insurance industry as a whole is upwards of USD80 billion per year. In addition to the cost of paying out on fraudulent claims that go undetected, there are significant costs associated with recovering fraudulently obtained insurance payouts through the courts—on average around USD10,000 per case.

If someone submits a suspect claim, the clock starts counting down—and the company’s investigators have a set amount of time to look into the claim before the company is obligated to pay out on the policy. In the past, it could take weeks for a team of ten people to comb through hundreds of online log files, claims documents and customer contact center records to uncover proof of fraud. In certain cases, the company simply did not have enough time to collect the evidence needed.

Customers increasingly prefer the digital channel to search, compare and buy their insurance—and fraud is also moving online. The company realized that identifying fraudulent behavior faster on the digital channel would give investigators a valuable head start, and looked for a way to enable the new discovery capabilities.

By reacting fast to detect and prevent payouts to fraudsters, we can avoid costs, reduce our exposure to risk and offer competitive premiums.

Spokesperson, leading insurer

Transformation story

Creating an event-driven approach to fraud prevention

To solve the challenge, the company deployed an automated, events-based alerting system across its digital sites—enabling it to use business rules to highlight potential instances of auto insurance fraud to its investigators, and to replay suspicious user sessions.

The insurer now has the ability to model common scenarios for fraud, and build triggers that pass the information to the right people in the business.

One common example is misrepresentation, when a person changes their details to find cheaper quotes. If a person initially states on the website that they that they are married with three children and then immediately afterwards generates an auto insurance quote for a single person with no dependents, the company can automatically prevent them from checking out the policy online, send a recording of the session to the contact center, and schedule a callback to validate their details before they purchase. Better still, the company can detect this type of fraud regardless of whether the person requests the first quote on a laptop and the second quote on a mobile device.

Automated alerts are now widely used across the business. In addition to the claims fraud department, the company’s security operations center uses the solution to detect digital attacks, and digital teams utilize the solution to identify and resolve customer experience issues.

The events-based monitoring solution has exploded in popularity, and the insurer now has a backlog of business requests for access to the platform. Monitoring capabilities are especially valuable to the IT support team, who can reduce the time required to test new business- and customer-facing services, and to resolve technical issues in the production environment.

In fact, the monitoring capabilities are so precise that the solution is uncovering technical issues on the company’s website servers faster than purpose-built scanning solutions. Maximizing availability in this way also delivers substantial commercial benefits, because less end-user downtime translates into fewer missed sales opportunities.

Results story

Staying one step ahead of fraudsters

The company’s claims fraud team has embraced automated alerts to reduce the time to build airtight cases. Investigating fraud is often a complex process, and any opportunity that the insurer has to reduce the time required to gather strong evidence can make the difference between being forced to pay out on a fraudulent claim and stopping it in its tracks.

For example, if someone files an accident claim on a vehicle with a garage address in a low-income area, but the location metadata on their mobile app shows their location is a high-income area hundreds of miles away, the company can quickly determine if they provided a false address to obtain a lower premium. On average, the company has significantly reduced the time to gather conclusive evidence of fraud.

The ability to store, replay and analyze digital sessions at any time helps investigators to prevent fraudulent claims from making it to court. If the insurer sees that a customer has been falsifying details across multiple quotes a couple of months before making a claim, it can call them, present the evidence and explain that it will show the session replay in court if they decide to pursue it. Every claim that the company prevents from going to court saves a substantial amount—and the company estimates that it avoids a great deal in legal and other associated costs every year.

Automation is a major asset in combating some of the more common types of fraud, particularly people attempting to re-insure vehicles that have been written off. By scanning for decommissioned vehicle identification numbers, the company can instantly put a hold on the affected policies and quickly contact the people involved.

Based on the positive impact of its event monitoring and session recording capabilities, the company is planning to extend the solution to its marketing teams. As consumer preferences shift towards shopping for policies on the digital channel, driving conversion online and on mobile is becoming increasingly important. The company plans to set up automated event alerts at each section of the purchasing funnel—enabling it to uncover the reasons for customers leaving the site prematurely, and find ways to create a more compelling experience that drives digital sales.

Criminals are using smarter methods all the time, and it’s essential that insurers stay one step ahead. By reacting fast to detect and prevent payouts to fraudsters, the company can avoid costs, reduce exposure to risk and offer competitive premiums to its customers.

Leading insurer

This leading insurer provides a full range of insurance and financial services.

Solution components

  • Acoustic Customer Analytics

Take the next step

To learn more about IBM Tealeaf solutions, please contact your IBM representative or IBM Business Partner, or visit the following website:

View more client stories or learn more about IBM Watson Customer Engagement