The Necessary and Growing Use of Data
In the use of large data sets, it's time to move beyond a focus only on analysis and consider the power of the process.
While a company may be primarily a claims management firm, in reality, it is a data management firm that handles claims. As with virtually every company today, data—and its analysis—is at the heart of everything you do, whether it’s being captured, validated, mined, translated, analyzed, shared, used for projections, or stored. The statement, “You’re only as good as your data,” becomes more valid every year as data sets become larger and algorithms evolve and are tweaked continually to deliver ever more refined and useful results.
According to ISACA, a nonprofit association of more than 95,000 information technology professionals worldwide, “Organizations with effective data analytics can make stronger business decisions by learning what is working and what is not, what they are doing well, and where they need improvement. By analyzing their information, enterprises can make decisions to help increase profitability, improve performance, and identify and manage risks.”
Large data sets and the analysis of that data touch most individuals every day, yet it is likely that many of us focus on the outcome of the analysis and do not consider the process. A commonplace example of the extraordinary power that data and formulas wield is the massive results that appear almost instantly from a Google search, results created by stunningly complex calculations applied against an enormous amount of data—a combination that has made the company a ubiquitous name, a verb, and a global powerhouse in less than 20 years.
Technology for collecting and managing data is sophisticated and often highly automated; we built the machines, and the machines do much of the work. But they don’t do all of it. One critical area where human training, experience, and judgment are essential is when we get to the application of results. We collect the data, apply algorithms, and output the results, but then decisions must be made on how the data is used. How are we using the analysis to help our enterprises? How does the data translate into actions to improve both service to clients and operations? What are the practical uses for managing claims? Here, we’ll look at a few of the applications and benefits of data analytics.
Operational efficiency can be defined as finding the optimal way for allocating resources at a given time and in a given situation. Analytics plays a key role in making such decisions, whether by supporting management decisions or as part of an automated activity.
This also leads to another key practical benefit—greater staff productivity. For example, consider a specialized claims adjusting group that focuses on large, complex, and technical losses. It can use its proprietary database in a number of ways to assist with claims. The database captures large amounts of data worldwide from a broad range of losses. Types of claims in the database can range from flooded electronic component factories in Thailand (property and business interruption claims) to online business interruptions due to distributed denial of service attacks against Internet servers to a professional indemnity claim against a software manufacturer for development of a custom application that was alleged not to perform as originally specified.
Using the database, claims coming in are assigned to the right adjusters based on analysis that combines characteristics of the client, claim, claimant, loss, location, adjuster load, adjuster specialty, and level of expertise with the objective of optimizing outcomes such as quality of service and reduced cost. Combining the data with retirement and age analysis statistics allows for a view on assignment volumes versus adjuster loading volumes, and assignment volumes per industry sector, product line, and peril codes versus adjuster age analysis in order to determine service areas where shortages of experienced adjusters may occur over time.
Overall, statistical analysis of historical data helps build predictive models that support operational decisions in areas such as assignments and future staffing needs. One particular type of data analysis takes into account activity data and reconstructs process flows, which helps identify opportunities for system-wide process improvements. The database also can measure and track claims volumes per country, region, industry sector, product line, and peril and then provide statistical forecasting on claims volumes.
Client and Product Insight
Data analytics is extremely useful in scrutinizing emerging client needs and helping to design product offerings to suit your clients’ specific requirements. In this type of analysis, data about past services is combined with industry trends and product level characteristics to identify opportunities for evolving products, developing complementary products, and optimizing product bundles to meet the nascent needs.
Such product design can be customized to a specific client based on a model built using data for other clients with similar characteristics. Fitting a product offering to a market’s needs has been done for as long as business has existed, but newly available data and technology allows for evidence-based models that complement expert intuition and, in some cases, uncover opportunities never considered before.
By the time a transaction is done and quality is measured, it is frequently too late to take any action to improve the quality metric of that particular transaction. Data analysis provides us with powerful foresight to ensure client expectations are met or exceeded for each individual case. In one typical situation, a rules-based software “engine” confirms data validity before a report is submitted to a client and ensures specific service-level agreement requirements are met by notifying the adjuster when some action needs to be taken.
For the claims file, diary entries are automatically generated for adjusters based on such rules. These rules are derived from evidence discovered in analyzing data, when patterns emerge from initial analysis and can then be used as templates for additional, more sophisticated analysis. In addition, predictive models can score various potential risks associated with the claim and notify the adjuster to mitigate the risks ahead of time.
In previous columns, I have mentioned my passion for strategic planning—it should be a mandatory activity for every C-level executive and for those who aspire to that level. Data analysis is the foundation for this type of planning. In addition to providing client and product insight, data analysis supports strategic planning with competitive analysis and market forecasting.
In the case of forecasting, historical data and trends are used to predict future demand in different countries and from different industries on near-term, midterm, and long-term horizons. The forecasts are used to make strategic decisions in relation to geographic and/or product expansion, acquisition, hiring, training, etc. Statistical forecast models also can be used for improved client stewardship, factoring in seasonal adjustments, and preparation for catastrophic events.
Companies have to make decisions regularly regarding investments in technology systems and the processes the technology supports. These systems can be very large, take years to implement, and affect every facet of operations, and consequently, companies become dependent on them. Technology investment decisions should allow you to capture and share data more quickly, accurately, and securely so you can improve the efficiency and flexibility of claims management and better serve clients.
As with any investment, there is the responsibility and pressure to invest wisely and to always consider the end results and how they benefit those within and outside of the company. Data analytics is a critical area where our investment has grown almost annually, but the return it has generated clearly has been a high multiple of the cost.
In our industry, the practical contributions of data analytics are beyond question. It is now a truism that good data is essential, and thoughtfully analyzed data and strategically applied results are significant drivers of success.