The Data Literacy Imperative - Part IV: Developing Data Literacy
Read - The Data Literacy Imperative - Part I: Building a Data Literacy Program
Read - The Data Literacy Imperative - Part II: The Data Literacy Body of Knowledge
Read - The Data Literacy Imperative - Part III: Data Literacy Assessment
This is the fourth and final article of the Data Literacy Imperative series. The series begins by describing the what, why, and how of building a data literacy program. Assessment and planning—key elements of the program—depend on a defined Data Literacy Body of Knowledge (DLBOK) that is described in Part II of the series. Part III describes how the DLBOK is used to conduct data literacy assessment. Now let’s examine the processes and practices needed to develop data literacy in individuals and in organizations. The process begins with assessment and planning, but those activities alone will not achieve your literacy goals. They are only the beginning. The full scope of the effort to grow data literacy is illustrated in figure 1.
Figure 1. Growing Data Literacy
At the macro level, growing data literacy is a four-step process:
Assessment is the essential first stage with the goal to know where you are.
Planning builds on assessment to decide where you want to go and how to get there.
Execution carries out the actions that are identified in the plan.
Measurement evaluates progress both in the execution of actions and achievement of goals.
Developing data literacy is not an event.
It is a journey that is guided by periodic assessment and continuous planning.
Each step is performed both for individuals and for organizations. Individual literacy is important to be sure that every person who works with data is able to create value and minimize risk. Organizational literacy is specifically managed because it is much more than the aggregate of individual literacies. Expect multiple iterations—perhaps sustained iteration—because developing literacy is not a big bang undertaking. It is not an event, but a journey that is guided by periodic reassessment and continuous planning.
The Human Dimension—Individuals and Organizations
As briefly mentioned above, developing data literacy is important both for individuals and organizations.
Individual Data Literacy. Individuals may be self-motivated to develop data literacy, or they may be driven by management initiatives. Every data worker needs to be data literate, though topics and levels of literacy vary among individuals. Data workers include anyone who collects, stores, manages, or analyzes data as part of their responsibilities, as well as anyone who uses data presented as charts, graphs, or reports to make decisions and take actions.
According to a Brookings Institute report, nearly every job has become a digital job that requires data skills. These are certainly technical jobs, but also lawyers, teachers, automotive technicians, nurses, office and clerical staff, and more. The report links digital skills with job resiliency—the ability to adjust to changing job requirements. With this in mind, it is important to consider who in your organization needs to have data skills and how to foster interest in data literacy.
Organizational Data Literacy. Organizational literacy is addressed from two perspectives: roles and groups. Roles are based on what a person does—business executive, business manager, business analyst, accountant, auditor, sales person, data analyst, data engineer, software engineer, data scientist, data architect, data steward, etc. Note both business and technical roles in the list of examples. For the purpose of managing data literacy, roles describe a collection of data workers with similar responsibilities and data interests.
Managing literacy by groups is based on how the organization is structured as business units, departments, functions, teams, etc. Literacy can be assessed, planned, and managed at several levels of organizational hierarchy based on organization chart structure. It may also be managed for cross-functional teams such as a data governance group.
When defining the scope of a data literacy program it is important to
identify the individuals, roles, and groups that are the focal points.
Identifying which individuals, roles, and groups to focus on is important when defining the scope of a data literacy program. For many organizations—especially large enterprises—it makes sense to start with the most essential individuals and organizations, then expand over time as the program matures.
Assessment
Assessment is the essential first step for developing data literacy. Determining the current state based on a DLBOK quantifies where you are, and it provides information needed to identify gaps and plans for growth. Individual assessment tests data knowledge of people to quantify data literacy as literacy scores. Scores are compared with target levels to identify gaps. For more about the assessment process, refer to the earlier article about assessment.
Organizational assessment aggregates individual scores by groups and roles, comparing them to group and role targets to identify gaps. When assessing at multiple levels of organization hierarchy, it works well to aggregate individual scores at the lowest level of interest and then rolls up to higher levels. Collecting assessment data in a multi-dimensional data structure (see the data model in figure 2) makes aggregation and roll-up easy, and supports slice-and-dice analysis of data literacy.
Figure 2. Data Literacy Measures as Dimensional Data
The fact table at the center of this data model contains the values for actual and target data literacy scores. Each row contains scores for one person, in one organization unit, in one role, for one DLBOK topic, and for one assessment date. Dimension tables contain the needed data about topics, persons, organization units, roles, and dates. The structure supports situations in which one person may work in multiple roles and/or organization units. It also supports multi-level organization hierarchy where one organization unit may contain several subordinate units. With OLAP technology, all aggregation and roll-up is automated, and the ability to see literacy levels by any combination of the five dimensions is fully supported.
Author’s Note: I intentionally included a data model to illustrate a point. Some readers—especially data modelers—quickly understand the model without explanation. Others need to read the explanatory text and study the model to understand. (If you want a 90-second introduction to reading data models check out this video.) If I had used a data pipeline diagram instead of a data model, that would be intuitive for data engineers but need an explanation for many others. A digital transformation strategy map would make sense to many business managers, but be less clear to people in more technical roles. This illustrates an important point about data literacy: Each of us has areas of knowledge that are intuitive and based on depth of experience, and we all have areas where we need to expand our knowledge. Broad knowledge across the entire DLBOK with depth of knowledge in the areas of your responsibilities is a realistic and practical data literacy goal. |
Individual Planning
The basics of planning are straightforward: know where you are, know where you want to go, and define the path to get there. These basics hold true for data literacy planning. Results of assessment are the starting point—knowing where you are. Next, define the destination, as data literacy goals. Then you can map out the journey from where you are to where you’re going.
Goal Setting is the process of prioritizing, refining, and sequencing data literacy targets. Targets are initially set for each topic as part of the assessment process (described in Part III of this series) and are used as part of gap analysis. At this point, you may want to refine targets depending on the number and size of gaps, and on the relative importance of each topic for your job roles and responsibilities. Identify the most urgent gaps and limit the scope to what is realistic in 6 to 12 months. Decide whether to close each gap, achieving the initial target set at assessment, or to narrow the gap by adjusting the initial target.
Learning Resources are the means to develop new knowledge and skills. Training resources are the obvious first choice, including in-person and online classes. Look for data literacy certification programs, commercially available classes, software vendor training, university continuing education programs, and more. Training classes are a good beginning but think beyond training to tap other learning opportunities. Are there others in your organization or among your colleagues who would be interested in group learning and discussion activities? Are there people able and willing to act as coaches or mentors? Are there current or upcoming projects that offer an opportunity to learn from experience?
Learning Plan development brings together goals and resources and plots them on a timeline—for example: (1) complete this class by this date to acquire this knowledge, (2) participate in this project for this time period to gain this experience, (3) work with this mentor for this time period to learn these skills. The learning plan is really your roadmap to goal achievement. It describes what you plan to do, when you plan to do it, and what you expect to accomplish.
Individual Execution
Nike says it well: Just do it! You’ve made the plan. Now undertake the study, engage with coaches and mentors, participate in discussions, get involved in projects. Take the actions that are defined in your plan.
Individual Measurement
Don’t wait until the end to measure. Evaluate study progress as you go. Are you completing classes on schedule as planned? Are you learning what you need to learn? When the answers to these questions are a definitive “yes” then you’re on the right track. When “yes” is not the answer, don’t hesitate to adjust the plan. You may want to change resources, change timelines, or, as a last resort, adjust goals.
Organizational Planning
Goal Setting for Capabilities considers the increased ability of groups and roles when data literacy is improved. Organizational planning is similar to individual planning: current state, goal setting, and mapping the path from start to finish. The assessment describes the current state, and some of the goals are similar with individual knowledge and skills translating into organizational capabilities. The aggregates of individual assessments quantify group and role capabilities, but that is where the similarities end. Aggregates of individual targets don’t directly correspond with group and role targets. Those targets are set independently as part of the assessment. Scope may be narrowed to a subset of gaps, a subset of groups, a subset of roles, or any combination of the three. Aggressive goals may seek to close gaps or less aggressively to narrow the gaps.
Goal Setting for Cultural Impact goes beyond DLBOK-based assessment to consider the impact of a data-skilled workforce and the business benefits of data literacy. Cultural goals include trust in data, use of data for decision making, integration of data into business processes, engagement with data and with self-service analytics, data and knowledge sharing, data-driven innovation, and advanced use cases such as prediction and automation. A pragmatic approach to cultural change starts by pursuing a small number of these goals. Don’t try to tackle everything at once. Choose the goals, then establish for each a benchmark against which progress can be compared. Some goals, such as trust, may need a survey-based benchmark. Benchmarking for others can be achieved by observing who is using data, how, and for what purposes. Set time-based goals but be realistic about timing. The time frame is different than for individual goals because organizational and cultural change isn’t readily achieved in 6 to 12 months.
Building literacy into HR practices, literacy as governance priority, and implementing data coaching practices are effective ways to drive organizational growth of data capabilities.
Literacy Growth Resources are the tools that are useful to elevate organizational data literacy and to drive cultural change. Incentives and motivators are the tools that influence individuals to actively pursue data literacy. These may be as casual as public recognition of data literacy skills and achievements, or as formal as building data literacy into job descriptions and performance reviews. Building data literacy into HR practices, making literacy a governance priority, implementing data coaching practices are effective ways to drive organizational growth. Organization, role, and process changes help to weave data into everyday business activities—perhaps a change such as embedding data stewards into selected business units, or charging department heads with the responsibility to provide specific data up the management hierarchy.
Literacy Growth Planning plots changes on a timeline. Define what incentives, motivators, and changes will be implemented and when. Also, know who will implement. Designate responsible organizations and people for each implementation.
Organizational Execution
Again, just do it. Execute the plan. Implement the incentives and motivators. Make the changes in organizations, roles, and processes.
Organizational Measurement
As described for individual measurement, it is important to measure as you go. Don’t wait until the end to know if what you’re doing is working. This is especially important when driving organizational and cultural change. Assessments and aggregation are a practical basis to measure group and role capabilities. Measurement of cultural progress uses the same basis as for establishing the benchmarks.
Cycles of Improvement
Data literacy is not a project. It is a lifestyle.
Growing data literacy is a continuous improvement process, both individually and organizationally. Proven practices of continuous improvement—iteration and feedback—are an important part of the data literacy journey. Continuous planning is more important than having a plan. Expect a cycle of plan, do, evaluate, and plan again. That is the inner loop of the diagram in figure 1 and a key to managing through one iteration of literacy growth. But data literacy is a big undertaking with broad reach across organizations and deep reach into business units and processes. Grow data literacy incrementally by taking on one manageable iteration at a time. Assess, execute, reassess, then broaden the scope for the next iteration. Ultimately, data literacy is not a project. It is a lifestyle.
A Commitment to Data Literacy
Every enterprise is data-driven—either intentionally or incidentally—because data is everywhere. Public or private, non-profit or commercial, and in every industry the people in your enterprise work with data every day. Data isn’t free; it is an investment. To optimize return on that investment, we must also invest in data literacy. From a Balanced Scorecard perspective, Kaplan and Norton show a causal chain that starts with learning and growth. What learning could be more timely, more pervasive, and more central to the data-driven enterprise than learning about data?