A typical auto claim process starts with a phone call from a policyholder who has just been in a car accident. The carrier representative collects details about the accident, and the claims processing system passes information to a claims adjuster’s queue. The claims adjuster then starts the investigation and may order incremental data that he or she thinks is most appropriate for each case. However, this process is labor-intensive. It may take 45 days or more to close the claim.
Now consider this scenario: an agent receives a phone call from a person who has just been in a car accident. The representative immediately sees all of the relevant data about all involved parties as it fills the screen through a data prefill product for quick validation and customer confirmation. The agent instantly confirms the person’s name, address, and vehicle identification number (VIN), as he or she collects details about the accident. Once the accident details are captured, the data is evaluated against an external database that indicates the claimant actually has coverage with multiple carriers. The claim is automatically directed to the carrier’s subrogation unit for further investigation.
Scenarios like this one are surprisingly uncommon. In areas of the business like personal lines, quoting, and underwriting, carriers have embraced data and analytics to improve profitability and reduce costs. Yet, they have not applied the same approach to claims—where the vast majority of a carrier’s premium dollars are spent.
Data and analytics can help carriers create a more efficient claims process: one that is both cost- and time-efficient, and that minimizes losses related to fraud while enhancing customer service. To reap the full benefits, insurance carriers must proactively supplement their internal policy-level data with external data.
Intuitively, carriers know the value of external data to the claims process, but they typically use it in a reactive manner. For instance, special investigation units (SIUs) order point-in-time information from police reports, medical reports, and public records data. However, by using external data reactively, carriers are leaving holes in the claims process—paying potentially fraudulent claims, missing subrogation opportunities, and allowing severe claims to escalate in the hands of inexperienced adjusters—while simple claims languish on adjusters’ desks, driving up handling costs and negatively impacting customer service.
In contrast, a proactive claims handling approach takes full advantage of multiple data sources and analytics engines. From first notice of loss (FNOL), carriers can evaluate the claim and route it to the most appropriate person or department: sending suspect claims to SIU, subrogation opportunities to investigators, potentially severe claims to experienced adjusters and fast-tracking low- or no-touch claims for payment.
It doesn’t stop there. After all, the claims process is dynamic, and the processes that support claims processing need to be equally dynamic. As claims are updated with new information, a proactive approach gives carriers a way to easily re-assess and re-route each claim, ensuring that each one is still in front of the right person at the right time.
Advanced analytics techniques can help a claims organization to develop more sophisticated, efficient ways to manage the claims process. Here are some commonly used advanced analytics techniques:
- Predictive models use algorithms to identify patterns in data. Claims are scored and then routed to the appropriate claims area, such as SIU or subrogation units.
- Business rules alert carriers when specific situations arise. These business rules can be customized according to a carrier’s particular needs.
- Identity matching monitors all claims for entities—people or VINs—that are on a watch list. Carriers are informed when the entity appears on a claim.
- Data search enables carriers to search structured and unstructured data for words, phrases and names. This tool is especially valuable, as unstructured data is notoriously difficult to parse.
- Relationship analytics find links between claims or claimants. This can be an immensely powerful tool for fraud investigations—for example, to detect links between claimants and known suspicious entities. When combined with external data, this can be particularly powerful by identifying relevant relationships that aren’t visible to a carrier using their data alone.
These tools can be even more powerful when used in combination with each other to take advantages of the strengths each tool independently brings. Identity matching and predictive model scores, for example, can be used to enhance business rules to reduce false positives and improve outcomes.
Pulling External Data
Carriers can reap the true benefits of data and analytics by augmenting their internal policy and claims data with external data sources.
“Insurers need to start leveraging the external data that is available in the marketplace,” advises Deb Smallwood, founder of Strategy Meets Action, an insurance-focused research and advisory firm. Smallwood recommends looking at the entire claims process to see how external data can bolster an insurance company’s internal data capabilities: “It helps to start at the beginning of the process,” she adds. “Upon submission of a claim, insurers can pre-screen for fraud. They can also use data prefill capabilities, for example, drawing from accident reports to help populate and validate information in the claim.”
Often, claimants are in a highly stressed state when they report an accident, which can adversely affect the accuracy and completion of the report. With insight from external data sources, customer service representatives can verify accident and policy information while offering more personalized customer service. More critically, the validated data results in greater accuracy of the claim details so the carrier can process the claim in the most efficient way.
With a validated claim, carriers can apply predictive analytics to fast-track low- or no-touch claims to payment, minimizing the time and cost of handling. They can also query external data sources for information about a claimant’s history of prior claims. Excessive history may merit further attention from SIU. Additionally, carriers can find out if a claimant has coverage from other carriers, unlocking possible subrogation opportunities.
In the area of fraud detection, external data and multiple analytics tools are a tremendous combination. “Fraud detection needs to get pushed into the earlier stages of the claims process,” says Smallwood. “You still need SIU to look at aggregate data for trends, behavior, and fraud rings; however, insurers need to start to apply analytics on the front-end for real-time fraud alerts.”
Predictive models can look for data patterns associated with fraudulent claims, notifying carriers of suspicious activity. Carriers can create business rules to isolate claims that meet certain criteria; for example, to find cars that have been reported stolen and discovered burned. Identify matching can monitor claims that reference a vendor or party of interest, and relationship analytics can find links between people and claims. Data search is a particularly valuable tool, enabling carriers to search through unstructured text—adjuster logs, emails, and incomplete policy administration forms to name a few sources—for medical terms, names, or phrases.
Furthermore, public records contain immense information that can help carriers handle claims more effectively. By proactively monitoring public records for financial distress, bankruptcies, liens, judgments, criminal records, and death records, carriers can uncover valuable information. It is not uncommon for carriers to discover, upon monitoring public records, that they have been paying disability claims to deceased claimants.
Finally, medical data is notoriously underused in the claims process, but it can be tremendously useful. Medical data can help carriers determine the possible severity of a claim and can help identify providers who consistently exaggerate or invent injuries. Carriers can apply relationship analytics to medical data to help identify fraudulent ring activity.
Tip of the Data Iceberg
Rather than applying a linear attitude to the claims process, carriers should adopt a dynamic, data-driven approach of continually evaluating and routing a claim as information is added. This next-generation approach to claims handling relies on proactively using external data throughout the claims process, and applying advanced analytics to route the claim to the most appropriate person or department.
The benefits of doing so are tremendous. By paying out simple claims quickly, carriers can minimize their handling costs while enhancing customer service. Moreover, by flagging and routing claims as they come in, carriers can be more confident that they are handling the claim in the most efficient way. Aside from the inherent cost savings, this can unlock subrogation opportunities and detect fraudulent activity while there is still time to act.
Smallwood concurs: “The industry is at the tip of the iceberg in terms of the opportunities for full use of data and analytics in the claims process. Right now, it offers a competitive advantage. Soon, it will become the norm. The insurers that get on board and start to use it now will be in a better position when the market picks up.”
Every carrier aims to handle claims quickly, fairly and accurately. They invest heavily to earn their premium dollars, spending the majority of their dollars in the claims process. By proactively integrating external data and using advanced analytics, carriers can reduce costs and ensure that they are making the most of their resources. This means getting the right information to the right person, at the right time.