5 Steps to a Winning Data Strategy

As discussed in the first blog of this series, data is the foundation for staying competitive in the swiftly evolving world of connected business ecosystems. More precisely, success depends on the ability to harness and use of data to drive intelligent actions on-the-fly. For many insurers, this means overcoming the challenges of legacy data systems, which requires a solid data strategy. 

In this blog, we’ll explore the key steps necessary for selecting and incorporating the right mix of technologies into your (Read: insurer’s) data strategy, as discussed during our recent P/C data modernization webinar. The goal is developing a strategy that helps you gain a modernized data infrastructure that can evolve rapidly and continuously to meet any new competitive opportunity or threat. 

Key Steps to a Modernized Data Infrastructure

building a winning modern data strategy infrastructure infographic design

Step 1: Map Your Current Environment
, Define Terms & Determine Business Goals 

Approach data the same way as any other asset: understanding what data you have, where it’s stored, how it moves through your organization, who uses it and why. To do so, create a data visualization, including a map of the pathways data takes as it enters, circulates, resides and, in some cases, exits your enterprise. For this phase, it doesn’t matter whether the technology within your environment is a legacy solution or a new one. What’s important is to create data visualization. 

In addition, it’s critical for the mapping exercise to ensure everyone in your organization has a consistent understanding of the various types of data, the repositories where it’s stored and the systems that use it. For example, do you have separate underwriting, claims, actuarial and financial data marts? Do these marts have different names for different business lines? Where does the data for these marts come from? What processes use the data from the marts? 

Finally, this step includes identifying the key business outcomes, which establishes signposts to help guide your data modernization journey and supply business users with concrete progress metrics. 

Step 2: Identify the Primary Components Your Platform Requires 

Next, determine the major components you’ll require to reach your transformed state. These components fall into four general classifications. 

Data Sources. Every modern data strategy includes four basic types of data sources. The mix of sources you choose should support the goals determined in Step 1. 

  • Internal structured data such as a policyholder’s name, address and policy number.
  • Internal unstructured data such as narrative details within a claim.
  • External structured data such as information in an ISO bureau circular or pressure readings from a policyholder’s plumbing sensors.
  • External unstructured data such as insights collected from studies, news items, weather patterns, and social media postings or even photographs taken by a drone. 

Data Hub/Data Lake. Known as either data hub or data lake, this is the modern repository where your data resides and the associated tools for working with your data. As mentioned in the first blog in this series, modern data hubs are frequently based on technologies such as Hadoop or MongoDB. You’ll require tools to extract, reconcile and transfer data from legacy systems into the data hub and other tools for integrating data with business systems, such as your core suite. You’ll also need to determine where the hub will reside, such as on the Amazon Web Services (AWS) cloud. 

Analytics/BI. Layered on top of your data hub will be the various technologies for data analysis and insights. With data hubs typically containing billions of data points, it’s humanly impossible to fully mine, analyze and understand the available data. Fortunately, applying AI-based analytics solutions can turn massive quantities of data into usable insights.  

Business Access. Just as it sounds, this is the interface layer for your business users. Specialized technology tools can help you build dashboards that make it easy for your business users to perform analytics modeling, obtain reports and complete other self-service tasks. 

Step 3: Include Capabilities Required for Delivering Continuous ROI 

As data modernizations are sizable initiatives that bring together stakeholders across the enterprise, success is a continuous process and not a specific endpoint. To deliver business value continuously, you should build the following characteristics into your data strategy, as required.  

Modularity. Organize your deliverables into smaller, logical capabilities that were determined in the mapping step. Then leverage agile development and a DevOps approach to deploy capabilities quickly and frequently within each module, continuously iterating to add improvements and harvest gains to apply to the next cycle. 

Repeatable Processes. Identify the processes within your data strategy and execution that can be repeated and then standardize these processes to speed up development, rollout and business utility. Throughout this phase, plan to work closely with the business process analysts to determine which processes are essential and which are no longer needed in the transformed state. 

Automation. Given the size and scope of data modernization, you’ll need tools and technologies that incorporate increasingly sophisticated automation capabilities. Make sure you consider the current automation maturity and the roadmap for each technology in your stack. 

Effective Support Model.  Data-driven operations over the long term require developing and implementing a support model that fits your enterprise. 

Step 4: Focus on the Top 3 Target Outcomes 

Although your data strategy includes nuances specific to your enterprise, the top three most important outcomes of any initiative include: 

Achieving Excellent Data Quality and Accuracy. High-quality, accurate data is fundamental to extracting real-time insights. Anything less sets your enterprise up for disappointment and, potentially, adverse business outcomes. 

Democratizing Data for Business Use. When data is democratized, everyone in your organization has access to it, meaning there are no gatekeepers that create bottlenecks to effectively gaining and using insights. 

Reducing Time to Market. With the ability to integrate analytics into workflows, such as pricing and underwriting, you can reduce time to market from the traditional industry expectation of months down to weeks, or even just a few days. 

Step 5: Build in the Hallmarks for Success 

Last but not least, we’ve noted that the following characteristics are common to those who achieve the highest levels of data modernization satisfaction. We recommend you include them in your data strategy: 

Team up with the Business. Strong relationships with key business stakeholders provide you with powerful champions for overcoming organizational inertia and keeping a data transformation initiative on track. Regularly delivering business-centric features and capabilities helps sustain enthusiasm. 

Establish Data Governance. Robust data governance is a critical predictor of favorable outcomes. Never compromise on this characteristic. 

Use Proven Approaches. Leveraging proven architectures, frameworks, methodologies and delivery processes not only reduce failure risk but also boost the likelihood of advantageous outcomes. 

Get the Right Technologies. As you identify and evaluate the individual tech components for constructing your data platform, take into account: 

  • Maturity of the solution and the ecosystem surrounding it. 
  • Availability of ample resources for understanding and modifying tools. 
  • Strength of the user community for collaboration and problem-solving. 
  • Ability to take an incremental development approach to enable continuous delivery and innovation. 

Insurers that incorporate the aforementioned steps into their data strategies will achieve better business results. The next blog in this series, Data Modernization: Tips from Pekin Insurance, talks about how the data transformation initiative at Pekin Insurance demonstrates the key steps in action.  

If you missed the initial blog in this series, Overcoming Legacy Data Hurdles, be sure to check it out. 

You can also view the webinar here: A P/C Insurance Data Modernization Journey: Learn from Pekin Insurance's Success. 

Previous: Overcoming ... Next: Data ...