The Value of Data: Do it Once, Do it Right!
As claims organizations embark upon their respective transformations, there is no time like the present to address the desired future business state and the data capture necessary to facilitate analytics. With the appropriate mix of internal and external data, predictive modeling can help insurers target referrals, assign claims at first report/notice of injury/loss; identify the most appropriate resources; alert case managers to significant deviations in treatment; decrease SIU referral time; and ultimately reduce claim duration and loss costs. From an internal data capture perspective, the following questions are important:       

• What data can currently be accessed electronically?
• What data is currently paper-based, but should be captured electronically?
• What additional data should be captured going forward?
• What performance metrics could be used differently to help run the business better? 
• What new or future performance metrics should be captured going forward?

These questions should be considered during pre-planning and integration design for a new claims management solution, which will typically be able to invoke business rules based on captured data and in part, further facilitate claims management capabilities. 

It is imperative to consider how data will be used while designing the future state operating model and claims business processes. The good news is claims organizations already have much of the data they need to begin building effective claim predictive models. Deloitte's model building experience to date suggests that up to three quarters of good predictive variables come from internal insurance company data. The remaining predictive variables come from a number of external data sources.

The key is to know which data and what combinations can drive outcomes, and to know this information as early as possible in the claims management process. Supplementing traditional claims and other internal data with new external data can help to provide more effective claims segmentation. The amount and quality of external data including zip code level, census block group level, and household level information continues to grow rapidly. From data about a claimant's lifestyle, to demographic and socio-economic data, external vendors and data aggregators are capturing much of what is now publicly available. 

By capturing a few more demonstrated data fields and integrating this information into models sooner rather than later, a model can provide greater predictive power by leveraging 50 to 100 variables taken from background information about the employee/claimant, the injury, external public databases, medical data and other external sources. In the end, predictive models can essentially act as decision support "eye glasses" for the claims adjuster's mind. 

One other consideration involves data quality and readiness. There is no such thing as perfect data, and spending too much time and resources to pursue 100 percent data quality perfection is misguided. The more important focus is to find the most applicable data and put it to work in the claim process as soon as possible. The saying still holds true, "80 percent of something is better than 100 percent of nothing."

Analytical Capabilities
In any future state, an insurance company leveraging advanced analytics may need to demonstrate the ability to apply statistical principles and repeatable processes to deliver high-impact claims predictive modeling solutions. 

For some organizations, this can initially require leveraging the experience of others who specialize in advanced analytics and predictive modeling. Over time, however, some insurers will develop a formalized data mining team, with defined job roles and responsibilities. Roles may include a data mining department leader responsible for the organization's most critical analytics initiatives, communication to the C-suite, and the ability to lead a cross-functional team; a development leader with the ability to translate business problems into model design specifications; credentialed statisticians for developing model inputs and evaluating multiple candidate models; and programmers to help bring the model scores and corresponding business reason codes to life. To the extent the organization has already implemented underwriting predictive models, the deep statistical experience of these resources can be leveraged to build out the claims predictive modeling team.

In addition to model development, the claims organization should establish the performance metrics and business intelligence tools necessary to measure the impact of the data mining and predictive analytics solutions. 

Once the claims models are fully integrated within the claims business process, performance metrics can be used to monitor the impact that data mining is having on the claims life cycle (for instance, SIU referral time, loss cost reduction, adjuster efficiency, re-assignment rates, and so forth).

Integration Scope & Planning
As is the case in most major technology initiatives, data design and implementation planning can be critical early elements to achieving the desired results. Data design and integration planning should be looked at multi-dimensionally to endeavor to provide a greater return on investment. 

Design should consider specific elements supporting an enhanced claims management solution and advanced analytics. Quite honestly, it is hard to consider one without the other in today's rapidly changing world. While the proverbial "hood is up" and exposing the claims engine, organizations should confirm parts needed to effectively leverage the power of data, facilitate core claim business processes and bring new analytical insights to claim handlers (identifying the right resource at the right time) are installed. These new insights should be considered early in the claim process, after the initial party contact, and ultimately throughout the full lifecycle of a claim. Critical data and analytical considerations should be at the forefront of discussions with the specific technology stakeholders, claims professionals and analytics/business intelligence leadership. If approached from a multi-purpose point of view, then claims data and integration planning should serve many business purposes. With the proper planning, organizations can minimize the risk of incomplete or fragmented data requirements or partial solutions, ultimately providing a greater return on investment for claims technology initiatives.

Analytics Continuum
As recently as three to four years ago, the notion of advanced claims analytics was largely conceptual in nature. Most claims organizations were considering internal analytics and performing core legacy technology replacement as separate and distinct business initiatives. Unfortunately, a number of companies weren't effectively approaching the opportunity as an integrated program or portfolio of projects. When contemplating analytics enablement, some were primarily using core claims or basic claims data as the primary input variables for establishing basic routing guidance and business rules. At the time, this helped these companies build basic analytic capabilities, but likely left them short of where they should be to take full advantage of predictive modeling.

Instead, organizations should consider incrementally building analytic capabilities while learning and adapting over time. While significantly improved over the past five years, a number of leading claims management solutions being implemented today can lack true predictive analytics capabilities of their own. Insurers should understand how advanced analytics can be integrated with these new solutions. Now is the time to think about core claims process enablement, more effective business metrics and how to leverage new core technology solutions to help analytics become part of the organization's DNA. The graphic displays an analytics capability continuum. When embarking on the journey to replace current claims management solutions, the topic of advanced analytics should be 'top of mind,' as well as an important part of the early planning and design phases. Regardless of when an insurance company actually implements advanced analytics in the claims transformation life cycle, it is imperative to help set the organization up for achieving the desired results. 

By fully vetting and understanding the analytics horsepower of the organization, data capture, and business metrics the organization should be capturing to support a desired/required claims future state, the company can lay a strong foundation for analytics into the future as it changes its core claims technology. If approached correctly and from a holistic perspective, the insurer can improve its planning and design efforts today, to help bring analytics along for the ride into the future.

This publication contains general information only and is based on the experiences and research of Deloitte practitioners. Deloitte is not, by means of this publication, rendering business, financial, investment, or other professional advice or services. This publication is not a substitute for such professional advice or services, nor should it be used as a basis for any decision or action that may affect your business. Before making any decision or taking any action that may affect your business, you should consult a qualified professional advisor. Deloitte, its affiliates, and related entities shall not be responsible for any loss sustained by any person who relies on this publication.

As used in this document, "Deloitte" means Deloitte Consulting LLP, a subsidiary of Deloitte LLP.  Please see www.deloitte.com/us/about for a detailed description of the legal structure of Deloitte LLP and its subsidiaries.

Copyright © 2011 Deloitte Development LLC, All rights reserved.