Insurance companies gather and generate a lot of data -- some would say they're drowning in it. If data is one of an insurer's most valuable assets, then why don't more companies holistically manage data like they do their investments, their real estate, human resources and the like? Where is the data management office? Who is the chief data management officer? Is there more of a focus on delivering data with technology than there is on delivering business intelligence? Business Intelligence should not be all about technology.
With some aspects of an insurance cycle that is either soft or softening, there should be a greater sense of urgency around the formulation and execution of a true enterprisewide information plan so that any market advantage, however slight, can be exploited. It is essential for insurers to truly understand the availability and breadth of data contained within the enterprise -- data that can take on all forms -- from electronic media to paper records. And don't underestimate the value of data on paper because if the data is valuable enough, the effort and cost to convert manual data to electronic data can be a high ROI initiative.
To begin to get one's organizational arms around this, an information inventory must be performed. This effort is non-trival and should be structured carefully to investigate the areas deemed to best match up with strategic and tactical objectives. For example, today's market dynamic requires most companies to be better underwriters or better claim handlers and to grow profitably. Precision pricing, customer retention, claims fraud detection, product cross-selling, producer recruitment and other activities must be performed in new and better ways to beat the competition. All of these challenges require more intelligent use of information but none can be achieved if the information isn't readily available. Every company should match these objectives with the information required to perform them and then devise a plan to find, gather and expeditiously aggregate and organize the necessary data spanning a number of years.
You might say, "We can do this already!" But can you? Do you have a rich array of data readily available for the past 10-20 years of policies, claims, premium billing payments, agent/producer sales, underwriting variables, D&B data, MVR and CLUE reports, quotes/hits/misses, call center records, regulatory/compliance data, bureau information, competitor data, customer survey data, etc. If so, where is the data and in what form? Can it be used for statistical analysis, data mining and predictive modeling? Is there a plan in place to continue to grow this repository with ongoing business data?
Companies are also drowning in external data, but often they don't know it. Once internal data is understood, a detailed review of available external vendor data should be conducted. It is common for different pockets of the business to purchase data for a variety of special projects and ongoing applications and it is often unusual for there to be a solid understanding of all the vendor data available throughout the company and how it can or cannot be used. This is data that companies already pay for so it seems only logical to understand more broadly how it can provide greater organizational value.
The focus of any enterprise information inventory should be on simplicity, practicality, and usability rather than elegance -- the value is on finding the data, organizing it and using it for high ROI projects that can help fund other more complex projects. Don't wait until the data is perfectly clean and organized because that is rarely achieved. Most data is fine for typical analytics and business intelligence and data in many forms can be put to good use -- whether for data mining or predictive modeling. The more people that can obtain value from the use of company data, the better. Think of this process as the creation of an investment portfolio by starting with small "sure thing" projects that will generate returns for use by the overall business intelligence effort. As each of these small projects earn a return and organizational credibility, bigger and bigger projects can be undertaken, all financed by the benefits realized from prior projects. The company can then declare that it is swimming, not drowning, in its data.
John Lucker is principal, Deloitte Consulting LLP, Practice Leader -- Advanced Quantitative Services (Data Mining & Predictive Modeling). He can be reached at [email protected], 860-543-7322.
On The Net