It may be an exaggeration to say you can't manage what you can't measure, but it's certain that if you can measure it, you can manage it better. That is the premise behind business intelligence (BI) -- the technology-driven identification, aggregation and analysis of data to support decision making and process optimization. BI can be applied selectively to any business problem, but it also enables the use of data to inform all business decisions, from front-line tactics to the highest-level strategic decisions.
An intelligent business is permeated at all levels by BI, competing within a new paradigm of objective decision making. With their vast stores of customer information and other data, insurers have an opportunity to excel at business intelligence, but they need to implement the thinking, technology and culture needed to support BI-driven competition. We spoke with the experts to find out what it takes to build an intelligent enterprise.
#1 Align BI Initiatives With Strategy
No insurer will become an intelligent business by accident. The transition to the new paradigm depends on executive leadership's explicit embrace of BI.
First, a company's senior leadership must acknowledge that information is itself an asset that must be managed strategically, insists Kelly Byrne, director, Business Intelligence Competency Center, Zurich in North America, part of Zurich-based Zurich Financial Services Group (US$70.2 billion in annual revenue). "Our executive team really sees the value of harvesting information and using it to gain deeper insights into where we are and where we want to go," Byrne relates.
Business leadership drives the cultural change toward valuing information as a strategic asset, which includes determining how data gets created, stored and distributed, and how documents and other content are habitually treated, Byrne explains. "Information assets can be cultivated to increase their value even when we are not simultaneously realizing value from them," he elaborates.
"The 'intelligent business' is not something that happens within IT, but rather at each decision point executed by all Zurich people, every day, while doing their work of serving our customers," Byrne continues. "When your executive leadership understands this, you are well on your way to having a truly intelligent business."
Byrne notes that Zurich North America uses SAP's (Waldorf, Germany) Business Objects as its main BI delivery mechanism. "We use all of the various reporting engines and dashboarding technology within Xcelsius," he says of SAP's dashboard design tool. "We also use SAS [Cary, N.C.] for modeling and heavy lifting on the analytics front."
#2 Create a 'Single Source of the Truth'
Building an intelligent business requires holistic thinking, according to Jamie Bisker, insurance leader, IBM. Carriers need to avoid being too problem- or line-of-business-specific if they want to succeed on a larger scale, he advises. "Silos are much harder to create with holistic thinking."
As a young company, London-based Torus Insurance (capital of approximately $1 billion) is unburdened by siloed legacy systems. But for that reason, the carrier's experience shows the competitive importance of BI. The challenge the company faces is one of extremely quick expansion and diversification across product lines and geographies, explains Zahir Petiwalla, the company's head of business planning and management information. Market cycles are becoming shorter, with longer soft-markets and shorter market peaks, he says. "In this environment, it pays to be better, faster and leaner, and strong BI capabilities are a key enabler."
Torus's BI strategy has been to build out foundational data capabilities, including tools and technology. "We have established a single, companywide operations function as well as a single operational data hub, information broker and data warehouse," Petiwalla relates. "The design facilitates getting to 'one source of the truth,' as all data and reporting for the company worldwide comes from a single source."
The availability of consistent data across the enterprise is an issue not just for competitive success but also for regulatory compliance, notes Jojy Mathew, enterprise information management strategy leader at Capgemini (New York). For example, he says, Solvency II requires financial, risk and customer information to be integrated, but many companies have failed to meet this requirement. "One of the first steps to becoming an intelligent business is to align key information domains and sponsors, such as the CFO, chief risk officer and chief marketing officer," Mathew asserts.
Mathew says many companies are looking to bring together business semantics, data governance and operationalization of data quality to align information enterprisewide. "One of the biggest problems across all sectors is as simple as you calling something 'apple' and my calling it 'pear,'" he reports. "You need to make sure, through management of business semantics, that an actuary in London is talking about the same thing, policy or person as an actuary in the U.S."
Zurich North America's BI strategy includes a common lexicon and tightly integrated information, the company's Byrne relates. "This means combining information from various systems and sources, and making sure all the proper business rules are reflected in the integration and that we have a common integration point, which is our data warehouse," he says. "From there the information can go out in multiple channels. But the power comes from the integration."
Achieving that integration requires a rigorous process of learning and reconciliation, Byrne stresses. "In the past, information was created in silos, so when you integrate you have to resolve the parochial conflicts of the silos into an enterprise definition," he explains. "That drives clarity and understanding in the business and also helps you improve quality."
#3 Identify the Data That Matters
When transcending from operationally siloed data to a unified enterprisewide paradigm, an insurer must identify what data matters from a tactical and strategic point of view, counsels Ray Desrochers, COO of HealthEdge, a Burlington, Mass.-based provider of software to health insurers. "You need to evaluate all the data sources you feel matter to you in order to understand what they mean and how they play together," he says. "You have to evaluate your constituent groups to understand what they need to be successful -- you can't make assumptions."
One of the most common problems in insurers' business intelligence efforts is that business users don't know what information is available, claims Capgemini's Mathew. "In most companies, IT has built something thinking that the business would come, but the business never came," he comments. "Companies need to make information readily available to the business through business catalogs."
The key to success is distilling information down to components that matter at the right level, according to Torus's Petiwalla. "Pivotal to this is having the data at the right level of granularity while having confidence in its quality," he says. "BI is not only about technology, but about people and process as well, so building a quality-focused, data-centric operational culture is as critical as the technology."
#4 Create a Knowledge Worker Culture
Buy-in to BI must be achieved across the organization, says Syed Haider, a consultant with X by 2 in Farmington Hills, Mich. "Insurers must drive a 'knowledge culture' in the enterprise that recognizes the increasing role of knowledge workers and intellectual capital in maintaining the business' competitive advantage," he adds. "They must recognize that the organization's business processes inevitably include a knowledge-management dimension, or else the full potential of the intelligent business will remain unrealized."
The success of BI initiatives depends on the business users understanding that BI is a tool for them to use, not a service provided by IT, insists John Sheffield, director of software development at Delta Dental of Virginia (Roanoke). The insurer began implementing San Francisco-based Pentaho's open-source BI software for claims processing in 2007. Sheffield says one of IT's top objectives was to encourage the business to move toward self-service.
"We are the enabler, providing information to our underwriting group, marketing and operations so that they can make the real difference," Sheffield says. "Within process-oriented organizations there's a tendency toward excessive dependence on IT -- 'We'll just rely on them to give us what we need.' If that's your outlook on BI, don't even bother."