How to Create a Data Strategy - Part l
In the information economy, data is the lifeblood of an organization. It flows through every business process—it shapes customer interactions, guides product design and development, and facilitates strategic plans and decisions. Without data, an organization today cannot function for long, and it can’t adapt quickly to changes in market forces.
Modern executives view data as a critical asset that must be managed, nurtured, enriched and delivered in a timely manner to all employees, customers, suppliers, partners, applications, and systems that need it. They work hard to eliminate data silos, data defects, and data redundancies that block the free-flow of useful information. They take ownership of data, entrusting a key business lieutenant to design a data strategy and oversee a data program.
The question many executives face today is how to create a compelling data strategy and program that advances the company’s key objectives and goals. There are two major components of a data strategy. The first is a process methodology that sequences the steps required to generate a strategy document. The second is a business-friendly, graphical framework that depicts the major components of a data strategy.
Eckerson Group uses a 10-step process to help companies craft a data strategy document. The process is straightforward. It aligns with the process we’ve used to help many companies evaluate and optimize their data and analytics environments. The ten steps are:
- Build awareness
- Assemble a team
- Educate the team
- Assess current state
- Develop a vision and goals
- Develop recommendations
- Develop a roadmap
- Develop a business case
- Prepare for change
- Execute the strategy
In the next article in this series, I'll describe each step in detail. But for now, I'll note that the two most critical steps are the ones most often overlooked: step 3—educate the team—and step 9—prepare for change.
Educate. Data is notoriously hard to corral and requires aligning many teams, technologies, and processes, each of which have their own organizational orbits that are difficult to alter. A half- or full-day of education gets team members up to speed about the key components of data strategy (see “ figure 1 below) and gives them a common vocabulary for discussing data issues and approaches.
Prepare. You can argue that education is an ingredient of a well-formed change management strategy, which is the second most overlooked step. Without a plan that aligns everyone’s interests in data and appeals to the “head, heart, and herd”—the essence of any change management program—a data strategy is doomed to fail. Unfortunately, in the race to implement the data strategy, many companies shortchange this critical step.
The second element of a data strategy is the content. A good data strategy defines the mission, vision, goals, and guiding principles of the data program, along with its key sponsors, stakeholders, and success criteria. The strategy also spells out the organization, team, partnerships, architecture, and technology required to execute the strategy.
Ideally, the data strategy is depicted graphically so that business users can quickly grasp and understand it. A simple picture should convey a basic framework that shows the business value, objectives, key components, and supporting initiatives of the data strategy. The graphical framework (and strategy document) should be free of technical jargon and use as few words as possible. Figure 1 shows a sample graphical framework.
Figure 1. Sample Data Strategy
Here is a brief explanation of the components of the data strategy depicted in figure 1.
Bulls-eye value. Data is at the center of this graphic—the bulls-eye of the strategy. It is surrounded by the primary business value that data delivers to the organization: reducing costs and streamlining operations on the top half, and growing revenues, customer loyalty, and satisfaction on the bottom half.
Objectives and functions. Moving outwards from the center, the two main objectives of a data program are to “run the business” and “grow the business”. Each objective aligns with the corresponding benefits in the inner circle and consists of four components. The “run the business” function uses reports, dashboards, automation, and workflow to optimize business processes and performance. The “grow the business” function uses analysis, exploration, prediction, and data monetization to generate new insights and innovations.
The two objectives typically serve different audiences and require different tools, architectures, and self-service capabilities. In many ways, the world of data is bifurcated into two separate camps that conflict if not properly organized and balanced. (See BI Power Struggle: A Strategy for Success). The ultimate goal of a data strategy is to reconcile multiple polarities—operations and innovation; centralization and decentralization, speed and standards; self service and governance; and agility and architecture.
Supporting initiatives. Surrounding the circle are four supporting initiatives. Starting from the top, a business-driven program led by a top executive (e.g., chief data officer) owns and executes the data strategy. To the left, a data governance program maintains the integrity, consistency, and quality of data assets. To the right, self service tailors data access and analytic functionality to individuals, roles, and groups based on their needs and permissions. And at the bottom, a data infrastructure and services manages the capture, storage, transformation, and security of data as it moves between and among applications, systems, and users, both inside and outside the organization.
The data strategy depicted in figure 1 is just an example, and maybe not even a good one. For example, organizational components are implied, not depicted. However, the best data strategy conforms to the needs and culture of the organization that creates it. And although it’s ideal to create a graphical data strategy, the nuts and bolts of a data program should be described in a strategy document generated from the ten-step methodology above.
Future articles in this series will describe the ten-step process and content framework in detail. Please provide feedback based on your experiences creating a data strategy.
Read - How to Create a Data Strategy: Part II – Ten Steps to Success