Context:
Founded in 1839, Gore Mutual is Canada’s oldest property and casualty insurer. Self-described as “prompted by care, not profit,” Gore operates as a ‘mutual’ company, whose sole purpose is to provide insurance coverage for its members and policyholders. It employs 350 people in offices in Cambridge, ON and Vancouver, BC, and works through an extensive network of broker partners to deliver responsive service to personal and commercial customers.
Gore Mutual prides itself on innovation, on the “exciting investments we’re making in people, technology, products, online solutions and claims service.” A key goal in this innovation agenda has been creation of a data platform that would speed and enhance the distribution of information from a variety of data sources, accessed via a variety of delivery mechanisms, to business users in underwriting, product development, claims, finance and broker areas. To build a platform that would unify its information landscape, the company began working with Information Builders (IB) over a decade ago to create a self-service BI and analytics environment featuring full enterprise integration of information sources and use of data quality tools. The outcome has been replacement of historical reporting tools, such as spreadsheets and paper reports, with easy access to a set of information portals that deliver appropriate business intelligence to various groups within the organization.
Gore Mutual’s efforts to build a data solution have been more journey than destination. The company began with extraction of data from back-end systems to a data store, from which web-based reports were developed for the claims department. However, as Jamie McDougall, VP of BI and Analytics at Gore Mutual, explained, “the data was better, but the environment was completely lacking in governance, and it only served a subset of our users….we only had a few super-users with the appropriate skillsets to manually generate reports, so other business users had to either rely on them for information or create their own Excel spreadsheets and pivot tables.”
Since then, Gore has extended its Information Builders-based technology platform to achieve enterprise-wide visibility into corporate data. Gore enabled distribution of information through the creation of a common data model for its policy management system, for legacy applications used by underwriting and claims, and other corporate data sources. Working with Information Builders’ Insurance Performance Foundation (IPF), Gore has implemented a BI solution that provides industry-oriented advanced analytics and a mature data model, along with insurance-specific strategy maps, dashboards, scorecards, reports, and dynamic alerts that are designed to address the needs of different roles and functional areas within the business. Data can be filtered, segmented to answer specific questions, presented in multiple different layers of granularity, and visualized in tables, graphs, sparklines, and other easy to interpret formats.
Establishment of its technology foundation has served as a critical component in Gore Mutual’s evolution to a mature user of digital information. Successful technology deployment, however, also requires consideration of people, process and technology – a requirements triad that is especially important in the area of data analytics. Today, data assets and their interpretation are widely viewed as the ‘new oil’; information insights may help an organization optimize operations, refine market approaches or create entirely new products and services to maintain or enhance competitive positioning. But data assets yield benefit for analytics users and their clients only when appropriate processes are in place to ensure responsible use, to ensure that data insights translate to effective action, and that intelligence delivers real business value. Enabled by the Information Builders platform, Gore Mutual has worked towards organizational maturity in a number of ways, evolving best practices in data usage and analysis that provide confidence in outcomes and which have helped the business to explore additional opportunity in its information assets. As McDougall noted, “The organization has had success and is maturing. We had a legacy system, but we now know how a number is created because we have done that work. We’ve matured our data assets and cleansed them, and we no longer have to have those conversations. As a more mature organization, we are now trying to figure out what the data is, and where does it go. What’s next is more empowering. We can be different, we can look at things we couldn’t look at before because we are further down the path.”
Self-service for business users
For many organizations, the key to operationalizing data lies in helping employees across all roles make better decisions based on data. But a central impediment to this approach has been individual capacity to access and understand the information that is available: data scientists are a scarce resource, information support services and IT departments are often overburdened with information requests, and other staff resources have not traditionally been equipped to access or analyze data. For Gore Mutual, a core goal was the extension of data analytics throughout the organization, while also ensuring staff were able to manage their own information needs. This has been achieved in two ways – through information portal design and in efforts to improve data literacy.
According to McDougall, this process begins at Gore with extensive consultation to assess business users’ information needs. When designing data portals for various groups, the BI group at Gore Mutual seeks to identify “what will improve your business, what will help you deliver better services, and what will help you better work with your broker partners?” Once the BI team understands source data requirements and other data consumer needs, and has landed on the right metrics, it will build an IB-powered dashboard or portal for delivery into the end user’s environment. To feed strategy and planning around data distribution, Gore has developed a process for consulting with business users, which has included support from Information Builders. Presenting as an independent, third-party resource, IB staff proved effective in interviews, in meetings with executives, supervisors, managers and other teams to ask about satisfaction with current information delivery and about future needs.
Once the BI team has met with the client, typically in larger portal engagements it will wireframe requirements and then take the wireframe back to users to discuss how they will interact with the portal – where the data will be placed and why it would look a particular way. In McDougall’s view, this process protects against “surprises” and helps the BI team understand that they are meeting the client’s expectations: “my BI dashboard and portal developers are not commercial underwriters” McDougall observed, “and they might not understand what metrics a commercial underwriter would use. [The wireframe approach] ensures there is a good exchange back and forth. It enables the conversation around what do you want as a primary function, but it also encourages discussion of what else may be provided. People may not know what to ask for, but we can interact to deliver better service internally.”
Role-based dashboard development
Gore Mutual uses WebFOCUS dashboards as its primary information distribution system. The company may look to deliver complex or simple reporting, including portals created with IB’s InfoApps toolkit, depending on the user’s role. For example, Gore’s data platform delivers high level business performance monitoring – what’s happening in revenue, what’s happening in claims, in claims counts or experience – based on the Performance Management Foundation (PMF), which is aimed at senior level management, though others may also access this output. Other reports are targeted at different groups: “Portals are intended to be role based,” McDougall explained, “targeting the data at the level appropriate to the consumer. PMF is performance level, general ledger, primarily financial data, while operational portals that are very specific to a business function deliver what those teams need at the time they need it. An individual responsible for broker management, for example, could see that this particular broker is growing, understand the drivers of that growth or the loss experience, to enable a meaningful conversation with the broker. It’s about enabling the conversations and allowing better business decisions.”
To manage end user expectations around information requirements, the Gore team relies on an iterative process, enabled by the IB platform’s capacity to scale data distribution. As McDougall explained, the BI team may start with a single page portal, add multiple pages over time, or modify the presentation based on feedback in an agile approach that supports innovation: “We have a proper development, QA production environment stood up for the information stack, so we have good process for iterative improvement. This is based on a good technology stack, built using IB’s exceptional toolkit, and we have delivered good success.”
Over time, and based on repeated requests, the BI group has developed some foresight into what information will be most useful. McDougall explained: “Historically, the self-serve questions have been rather ad hoc queries – my broker partner needs to know all the policies in the book that have this certain characteristic, for example. But so many of these ad hoc questions have been repeated so often, that we have pre-built those queries and codified them in a WebFOCUS portal so that the user can go in, click the broker, identify the report they want, run and get the data they need, in order to deliver value to their broker right away.”
Education on data analytics – “We’ve come a long way, but we’ve got a long way to go”
Ensuring that the business user understands the data that is delivered, and how it should be applied, represents a different challenge. A data literacy gap may mean that individual staff members do not comprehend the complexity of how data is created and hence are ill equipped to make good decisions based on report output. Managers, leaders, and supervisors in particular have to be able to understand and draw data from diverse sources in their decision making, McDougall added, but it is often the case that a management individual has attained a leadership role based on technical competencies or skill in a specific area, rather than through his/her understanding of data – and may therefore be less committed to a data-driven approach to decision making.
To address this issue, Gore Mutual worked with Information Builders’ Professional Services Customer Education group on the design portions of various projects (to optimize data presentation), involving business users to help them understand what is available through the portals – and to encourage them to access new online views that provide more information and better context, instead of relying on a legacy report.
According to McDougall, Gore has made considerable progress on this front: “Brokers struggled in the early days of more complex pricing algorithms, and they were frustrated by the fact that they couldn’t understand how the price was being generated – they felt that their value proposition lay in helping the customer understand exactly how pricing was calculated. But over time, brokers have come to understand that they need to help the consumer understand broadly what goes into characteristics of risk, but they don’t necessarily need to calculate the rate for the consumer. They can work with the consumer and still add a ton of value without calculating the rate.” With more time, he expects that the maturity of internal staff with respect to BI and analytics will only develop. Many individuals have embraced the tools provided by the BI team, recognizing that these offer more information, and can add more value to staff decisions, to their broker partners, to their underwriting partners. McDougall believes that as these “front runners” succeed with the application of data, they will motivate their peers to follow, helping in the adoption process.
Ultimately, Gore views end user education as an ongoing change management challenge: according to McDougall, “we will always have people in our business who we will continue to have to work with to improve their literacy or their numeracy with respect to applying data to make the best business decisions. This is true in insurance, and other highly technical areas.” Recognizing that data skill requirements will only become more complex, Gore views improving data literacy as an ongoing exercise and is considering implementing competency training, data and analytics courses, and certification concepts.
Ethical management – a hot button issue
Based on the systematic computational analysis of information, analytics finds patterns in data that can highlight trends that are useful in a business context. But there is a darker side to analytics when this is not practiced in an ethical manner. In the identification of certain characteristics, analytics and advanced forms such as AI and ML in particular have potential to marginalize and ultimately disadvantage certain groups. Bias in analytics has recently begun to receive more scrutiny – and rightly so – as correlations based on imprecise assumptions can have significant impact on individuals and groups. This risk is apparent in analytics initiatives that rely on personal data; for example, the insurance industry works with many kinds of data but often focuses on demographics and other personal attributes associated with risk.
With proper oversight, though, it is possible to mitigate these risks. Gore Mutual utilizes items that are reflective of risk, but manages what data is employed in building an algorithm through policy that defines what is acceptable to layer into the analysis. In automobile insurance, where the majority of rating work is done, the categories of information that are used are approved by regulators. On the property side of the business, McDougall explained, the company would not consider certain elements – hypothetically, augmenting data with regional ethnicity based on geography – and would never incorporate that into an assessment. “We have to make sure we have acceptable practices,” he added, “we have integrity guidelines that appeal to different stakeholders as a corporate value which underline our business.” And it is possible to effect this kind of control and governance, when analytics are built on a mature data platform. According to McDougall, through data management achieved in work with IB, the data that resides inside Gore’s analytics environment is well contained via the detailed mapping processes. For governance requirements around other, ancillary data, the company is considering solutions developed by specialists in governance, such as IB partner Collibra.
Adoption metrics to drive usage
From an ROI perspective, user adoption of a new technology is an important metric: software that is not used is not a good investment for the business. Underutilization means that that the organization will not achieve the range of soft productivity or hard, bottom line results that it expected to see when it justified the acquisition of the software. Best practice requires, therefore, that the organization have insight into use of its software tools. Drawing on capability in the IB solution, Gore Mutual is able to capture the activity that is occurring within its analytics platform, and has built a separate portal for monitoring data that demonstrates usage patterns – which articulates what portal pages are being used, when they are being used, what is not used, and who’s not using it. “Most of our portals are delivered to specific consumers,” McDougal explained. “specific to business development, to claims, or to commercial underwriting, for example, so if there is there is a lack of use in one of those portals overall, we would have an idea of the functional area that is not using analytics. And we may have identification of individuals who are not using it.” Armed with this detailed information, the BI team is able to confer with users and remediate usage-related issues.
Moving forward and giving back, the benefits of mature data usage
In assessing the impact of its analytics platform, Gore Mutual can count many operational benefits. The company has built dashboards which deliver productivity information that has enhanced underwriting, revenue, and claims forecasting by empowering users to assess performance relative to predefined goals and to drill into claims and other data to uncover operational details. An overall view of company performance is provided through the PMF, which tracks achievement against financial targets, with comparisons on monthly, quarterly, or yearly basis designed to identify trends over time. These capabilities allow Gore to fine tune-operations with an ultimate goal of speeding time to value.
But how is this achieved, and what broader social value is achieved through mature, responsible data usage? A fundamental outcome of using clean data subject to ethical norms is greater accuracy, an essential foundation for sound business decision making. McDougall pointed to several other outcomes that demonstrate how the Gore’s data platform advantages employees, business partners, end clients, whose greater success ultimately returns to the company.
On relations with broker partners, he noted “For Gore, our brokers are critical business partners and there’s no question that having a better interface with them, a more informed business context for the broker, allows for improved management of our business and better working relationships with business partners. They are all independent businesses and we treat them, in many ways, like our customers.” Brokers, he added, sell multiple insurers’ products, which means that Gore must compete with other insurance carriers. For its end user customers, the broker offers an important value proposition by identifying carriers that are operating well, those with exceptional results, firms who takes great care of their customers during claims time, and by adding information on issues like the direction of the insurer’s business and whether the organization have a good, positive, social impact. Information sharing with brokers spells marketing benefit for Gore, improved insights, and better service delivery for clients.
For the end-consumer of insurance services, more efficient and responsible data provided by the insurance carrier will delivery additional value in areas such as claims response. For example, when a fire breaks out in an area, Gore’s BI and analytics tools and team can more quickly and reliably identify where this event is occurring, what policies are in place relative to the fire location, and contact the appropriate broker and the consumer to see whether or not they have experienced damage. “We have in previous loss occurrences, proactively contacted people that we knew were in flood areas, or in the fire location,” McDougall explained. “It enables better responsiveness – it’s about delivery of exceptional claims care and delivering quality.” And while brokers or end consumers of insurance services may not understand what is driving this responsiveness, McDougall believes efforts to develop staff facility with data technologies and analysis will “enable better processes, and better enable what we do as a business” so that Gore receives improved satisfaction back from clients.
For Gore Mutual, the evolution of an increasing sophisticated analytics capability has resulted in another outcome that provides new benefit to consumers. Using advanced data modeling capabilities, the company is now able to offer new products that address contemporary challenges. Climate change, for example, has resulted in more frequent occurrence of extreme weather events, including flooding and fires. According to McDougall, risk exposure associated with these kinds of occurrences has historically been very difficult to assess. However, using advanced modeling tools, Gore has been able to build products that can deliver client coverage while minimizing organizational risk. “Our job is to take risk and provide protection. We need to assess what level of risk is involved and have pricing commensurate with risk – and we can better assess your risk if we have better tools and data. The reason flood insurance was not available in Canada for so long is because insurers couldn’t effectively measure the risk – the only people that would have purchased overland flood in the past lived in very high hazard flood area. No one else would buy it. So you would have anti-selection (the people who bought it were the people who needed it), meaning you can’t spread risk. But that’s what insurance does, it pools many characteristics of risk and distributes it. Better modeling has enabled the better spread of risk.”
BI and analytics have helped Gore Mutual in many ways – the company has achieved improved quality of pricing, better information distribution for management, enhanced operational management for staff, solid performance, and good customer satisfaction feedback. Technology has served as the basis for this success, providing both technical analytics capabilities and the groundwork for the evolution of mature data usage practices, which in turn are enabling new opportunities. For Gore, the voyage continues.
