Skip to main content
Using decision management to transform claims

Karen Pauli of Tower Group recently wrote Technology Direction in US P&C Insurance Claims Operations: Transforming a People Business (subscription required). This is a great paper and highly recommended for those of you thinking about how to improve your claims process as well as those just thinking about how you can use technology to improve what you might have historically considered a manual process. It struck me that enterprise decision management (EDM) can really make a difference.

TowerClaimsThe first point Karen makes is that claims is, and very much considers itself a "people" business. Historically claims organizations have feared that relying on technology will allow fraud to slip by and in over/under payment. They also worry that technology might get between the claims adjustor and the customer causing a loss in the conneciton they value. That said a number of drivers have combined to force claims organizations to make more willing to consider technology solutions. First there is a huge risk from the impending loss of claims expertise due to the retirement of the claims adjusters of the baby boom generation (something we have discussed before in EDM Could Be a Fix for the Aging Insurer Population). This know how cannot be allowed to just leave, nor can it easily be replaced. The second issue is that of compliance with regulations like Sarbanes-Oxley. Demonstrating compliance with regulations can be very difficult if the process is completely manual. Lastly there is a driver from customer expectations. As shown in the graphic on the right, customers are used to the kind of service they get from leaders in the new economy. They are used to service that is seamless and 24x7 365 days a year just like eBay or the credit card company. Delivering this kind of service forces more and better self-service as well as consistency across channels.

So these trends are forcing automation on claims organizations, despite their reluctance.

Karen argues that, unlike other carrier segments (such as underwriting), in which leading-edge technology resulted in completely automated processes, the greatest benefit to claims operations will be in decision support. I am not 100% with her on this but let's continue. Karen goes on to divide claims processing into 3 subsegments in each of which I see a value for decision management (not just decision support).

  1. Straight Through Processing where payment can be made immediately
    Decision management of the claims payment decision and of the actions to take is critical to STP
  2. Fast Tracking some simple claims where additional information is required
    Decision management can help decide what additional information is required, the sequence, how to get it etc. and can then handle the response when it is received
  3. Referred for manual processing
    While this thread has a focus on decision support (to help the claims adjusters), there is a need for decision management around the decision to refer and why e.g. the potential for fraud or litigation

As Karen notes, the bottom line is that human skills are expensive and increasingly rare so you need to sure they will make a difference in the outcome before using them. Decision management can really make a difference in applying them less often and more usefully. The paper goes on to discuss various technologies for improving the claims process. Karen talks about four items:

  • Business Rules
  • Business Intelligence - including reporting and analytics
  • Workflow
  • Web Services

While this is a good list, I would talk about these slightly differently both because the use of rules and analytics together to make decisions has proven itself (particularly in insurance) and because her topic of "Business Intelligence" includes very disparate uses of data insight - both reporting/decision support and insight for use in rules. My list would therefore be:

  • Decision Management
    Business Rules and Predictive Analytics combined to automate decisions like whether to pay a claim, what additional information is required before a claim can be paid, which claims are at risk of fraud etc.
  • Decision Support (BI and analytics)
    To help those included in the process, when a person is needed. This might be reporting, predictive reporting etc.
  • Workflow
  • Web Services
    Including the use of Decision Services

Karen ends with a strong emphasis on predictive analytics to drive superior results in terms of predicting risk and fraud, potential for subrogation and need for reserves etc. This is something Tower Group has discussed before. I completely agree that claims is an area where predictive analytics (check out the FAQ here) can make a big difference. One thing to bear in mind is that compliance in predictive analytics an issue - can you show how your predictive model came up with its result. The use of predictive scorecards in insurance and the drive to make the results of models like neural nets less opaque. The need to use predictive analytics comes up regularly, for instance in Insurers Not Using "State-of-the-Art" Analytics Face 'Grim Prognosis' and Predictive Analytics a must for Healthcare Payers.

If you are looking for an overview of how decision management can help insurance you could try this presentation - Live from InterACT: Putting Enterprise Decision Management to Work in Insurance and the general Insurance section of the blog. We have blogged before about Insurers finding millions of dollars in avoidable claims and about using EDM to achieve Straight Through Processing in Insurance. For a discussion of the difference between decision support and decision management, check out this interview.

Technorati Tags: , , , , , , , , , , , , , , , , , , ,

related posts