How to adopt a business-driven data and analytics strategy
To say data is a critical driver of competitive advantage and corporate success is an understatement.
The ability to be nimble and compete on intelligence is one of the biggest competitive challenges executives face today. Yet a large number of companies still pursue an IT-led data strategy that doesn’t unleash the real value of data to meet business needs for today and tomorrow.
No longer a pure IT project, data strategy is now driven by business requirements and enabled by IT. To succeed, companies must shift to an innovative data and analytics strategy that allows data to be transformed into actionable intelligence with an emphasis on business value. This changes the focus to identifying the right data that is critical to meeting organizational requirements for effective planning and decision making.
Drowning in data
Technology has provided an explosion of data that can lead to unprecedented opportunities. Applications of artificial intelligence (AI), machine learning and deep learning rely heavily on data, making it a key driver of competitive advantage. But more often than not, valuable corporate data is trapped in silos or disparate data sources, or its usefulness is impacted by poor data quality.
An even more pressing problem lies in not knowing what data are available today and what data might be needed tomorrow to meet business requirements. Massive amounts of data and increasingly large data sets provide no value unless that data can be standardized, categorized, harnessed and applied to specific needs of the organization.
Since most digital transformation projects have an underlying data component, this is a critical challenge that executives must address to be successful. A strong data and analytics strategy can ensure companies have a flexible and effective data repository to meet current and future needs.
A step-by-step journey to success
The transition to a business-driven data and analytics strategy should take a phased approach to ensure that all critical elements are addressed to achieve the desired outcome. This approach includes several crucial steps, each of which is designed to inform and provide a foundation for the next.
Engage with business leaders early in the process to understand and outline requirements around as-is and to-be reports, KPIs, advanced analytical models/AI applications, campaigns, etc. These discussions should occur with the business stakeholders and their teams.
Once the as-is and to-be requirements are laid out and prioritized (e.g., must have, should have and nice to have) based on business needs and goals, break down the requirements into specific data elements.
Conduct a data gap analysis to determine what prioritized data are currently available and what data needs to be captured. This step is important to ensure that you don’t have a patchwork data program.
Conduct a traceability analysis and develop metrics to determine exactly where data are being stored. From here, a new architecture of data flow can be developed that identifies how the data are going to be gathered from disparate data sources and moved into a data repository. The final data repository should be effective and flexible, designed in a way that allows future data to be added as needed.
This architecture should be reviewed and approved by business leadership and IT to ensure buy-in.
Target the must-have and should-have business requirements and the data that would enable those. If there is a gap in data from what is required vs. what is available, identify a process to capture the additional prioritized data into the system
Create a Business Requirement Document (BRD) that captures the superset of as-is and to-be prioritized business requirements, data elements that would enable them, data gap analysis, data flow, latency and architecture, RACI & Traceability matrices, a logical data model and a physical data model implementation roadmap. These documents need to be signed off by key stakeholders from both Business and IT before proceeding to the implementation phase.
A business-driven data and analytics strategy ensures that valuable corporate data will be easily accessible for use in intelligence and data-driven decision making. It also is key to gaining the greatest benefit from data-dependent technologies such as AI and machine learning.
Using a phased approach helps ensure a strong foundation for future success, with critical proactive planning, needs and requirements assessment, and crucial buy-in from business leadership and IT.