A survey conducted by the data management platform provider Cloudera between July and September 2021 found that IT decision makers overwhelmingly consider enterprise data strategy key to ensuring business resilience—91% agree. Maintaining that resilience, and scaling enterprise data operations to meet the competitive demands of the future, takes an open and agile posture.
Why Is Flexibility So Important?
What Makes A Data Strategy Flexible?
So what does flexibility look like? Look for these characteristics, which are strong indicators that a strategy will promote agility—and empower rather than impede the necessary pivots to come:
How Can Organizations Become More Flexible?
“There's three components of the strategy,” said Duby, explaining how leaders can infuse their own organizations with greater flexibility. “One is the people on your team. Another is the process, which is the fit into the overall picture of the organization. And the third is the technology,” Duby said. “And they all have to come together in order to be successful.”
Click To Learn How To Take Action
Technology
Select technology with the potential to deliver ROI in early stages of product development that can scale to meet larger challenges as the need arises, rather than requiring developers to start from scratch after the pilot stage.
Prioritize investments in tools that can change and grow with your organization, and retire those which cannot. Resist the urge to invest in a new tech with one or few demonstrable use cases. This not only inflates tech expenses but can tie up valuable IT resources on the indefinite care and feeding of orphaned legacy systems. “A lot of times engineers get very excited about a technical solution but fail to look at how that technical solution fits into the overall structure, culture and processes of the organization,” Duby said.
Carolyn Duby
Field Chief Technology Officer, Cloudera
Organizations have to be able to take advantage of their data in the quickest, most innovative way…[so] you're able to spring back or change course to keep yourself competitive.”
Hear More About Flexible Resilience
Scalable
Your approach to data should be flexible enough to scale rapidly in either direction. Starting small with a clear roadmap to grow to suit a broader range of data or increased analytical demands is a common approach. But to be truly flexible, the enterprise data strategy should include offramps to shrink reliance on a data source or an analytical engine when they outlive their usefulness.
Regardless of an organization’s unique constraints and needs, a flexible data strategy can help extract insight from any data within its control or influence, whether structured or unstructured, in motion or at rest.
Expanding data sources and a growing demand for data-driven insights mean that a strategy focused only on present goals will soon restrain progress. Baking flexibility into your data strategy ensures that your organization will be ready to meet new needs as they emerge.
Goal-Specific
Flexibility isn’t measured around meeting artificial or tech-oriented benchmarks. Rather, the strategy must help an organization be more agile in meaningful and demonstrable ways. For one company, that may mean increasing efficiency by 17% by streamlining a previously cumbersome task; for another, it may mean enabling a 3-D view of business-critical information by connecting analytical tools to vital pockets of data, wherever they reside.
Consistent
The data strategy should present users with tools and reporting that are consistent and familiar between assets and over time. Employees and partners should not need to learn new tools and interfaces for each new data-driven application or project. Common metadata, nomenclature and documented procedures also allow users to access and manipulate information without affecting the source data.
Hybrid
The strategy should span on-premise, public cloud and private cloud options, with roadmaps to incorporate new data and application delivery. This includes the ability to evaluate the tradeoffs of each tech option on dollar cost and compliance implications, and the freedom to move workloads between hybrid resources. “You might repatriate something in the cloud back to on-premise and open a new data center,” Duby said.
Unified
Although hybrid resources are at work under the surface, from a developer perspective the organization’s data strategy should rely on a unified platform. This enables decision-making based on a comprehensive view of data across a wide range of sources—and accounting for both real-time and batch data.
Resilient
The enterprise data strategy should include comprehensive accounting for disaster recovery and operational continuity. “You need to know your requirements for coming back online, and to know the impact to the business if you have sudden downtime from a provider,” said Cindy Maike, Cloudera vice president of business and product solutions.
To achieve a maximally flexible data strategy, leaders should agree upon an organization-wide approach that addresses all three pillars—and begin taking incremental steps to put it into action.
People
Any successful corporate strategy starts with culture. Data tools mean nothing without proper education for the people expected to use them, an organization-wide understanding of how they’ll help the business achieve its goals and an openness to innovation and experimentation. Leaders must actively champion this acceptance and encourage a culture of change. For example, ensure your data strategy allows for projects to start small and fail without punishment.
Involving a wide audience of employees across the organization in developing and stress testing your strategy also promotes greater flexibility. These cross-functional teams should include business leaders, data experts, developers and creative designers. This ensures data-driven knowledge and skills become more commonplace, and the diverse audience can help identify impending pivots or trends before they become crises.
Process
Your data strategy should suit times of growth and contraction, and be documented in plain language that can be adapted as new lines of business are spun up, acquired or divested. “You need to evolve with changes in the business climate. Gosh knows, it's definitely changing now,” Maike said.
And it doesn’t have to tackle 100 unique problems in order to be valuable. Start with a small number of business questions, and build tools to answer those questions, rather than trying to extract value from every data point on record. For example, “a manufacturer might start by addressing one small quality problem,” Maike said.
Working with cloud partners can help you achieve this kind of flexibility, but that doesn’t mean surrendering oversight. Evaluate how their data security and data management practices support your process. “You need to understand the locations where your data is going in order to truly have ownership of it,” Maike said.