Build a data literacy program to fit company needs

Start with a baseline assessment

Before designing any program, leaders must understand current data capabilities to avoid overtraining experienced staff or overwhelming employees who need foundational training. A lightweight literacy audit helps reveal these differences early, grouping employees into four levels of data literacy: basic, working, tactical and strategic.

Skipping this diagnostic step often results in leaders adopting training pitched at the wrong level -- courses that are too basic for experienced staff but too advanced for teams that need foundational skills. Many organizations focus exclusively on basic-level introductory content when teams would benefit more from working-level literacy.

An effective assessment has three components. Once complete, organizations should appoint a program owner with enough organizational credibility and data expertise to steer adoption.

1. A workforce survey

Give employees a structured set of questions to measure comfort with data tasks tied to their roles and tools. A survey is more useful when it asks about specific tasks rather than abstract concepts. For example, "How comfortable are you reading a bar chart?" yields more meaningful insight than "How would you rate your data literacy?"

Self-assessments have limits because employees might overestimate skills they use infrequently. However, they quickly surface areas of confidence and uncertainty. This is especially helpful when comparing skill levels across departments and teams, which often vary more than expected.

2. Role-based capability mapping

Don't treat the workforce as a single unit. Identify two or three representative role groups and determine which data tasks each performs. Different roles interact with data at different frequencies and complexity levels. Mapping required capabilities to actual behavior reveals where skill gaps exist and helps target training more effectively.

For example, a marketing analyst, an operations manager and a finance director all interpret and act on data, but in different contexts. As such, each needs to develop distinct skills.

3. Discoverability and reuse

Many organizations create new reports or dashboards rather than reusing tested assets. Working-level data literacy includes knowing how to find, validate and adopt existing assets, and understanding why reuse matters. Duplicate reports often diverge over time, introducing inconsistencies that weaken data-informed decisions.

Difficulty discovering assets often stems from poor information exchange and knowledge sharing. Establishing conventions for analytics documentation, naming and distribution helps reduce rework and confusion. Teams that develop a common language for their data assets find each other's work more easily.

Not all skill gaps deserve equal attention

Identifying gaps is only half the task. Prioritizing which ones to close first requires connecting data literacy to business objectives. Be clear about what the organization wants to accomplish within a realistic time frame rather than aiming for an aspirational future state.

Start by addressing gaps that directly affect current objectives and offer the clearest ROI. Identify two or three business problems that improved data literacy would help solve and anchor the program to concrete operational or strategic outcomes. Early wins build the case for broader investment.

Other gaps matter more as the organization's ambitions grow. Teams need a baseline level of AI literacy and an understanding of how data supports AI systems. This isn't about knowing how large language models work. Rather, staff must know how to share data with models responsibly and evaluate outputs -- for example, recognizing recommendations that seem plausible but reflect a misunderstood prompt, or treating a confidence score as a probability rather than a certainty.

Align the program to your organizational context

Once gap analysis is complete, design a data literacy program for the specific context in which it will run. Three dimensions matter most.

Dimension 1: Industry and regulatory environments

Organizations in highly regulated sectors, such as healthcare or finance, already operate under strict accuracy and governance requirements. These obligations create constraints and advantages for a data literacy initiative. Compliance is a compelling rationale for investment, but programs must still address specific regulatory expectations alongside general analytical skills.

Dimension 2: Workforce composition

A single uniform program serves no one. A mixed workforce of knowledge workers, frontline operational staff and technical specialists needs differentiated pathways:

• Knowledge workers making planning decisions need to work on tactical literacy.
• Operational staff must understand the KPIs guiding their workflows and recognize when numbers appear incorrect.
• Technical specialists benefit from developing their data storytelling and visualization skills to communicate findings clearly.

Dimension 3: Corporate culture

Assess the corporate culture realistically. A program that matches the organization's readiness is key. If senior leaders still work on instinct rather than data-informed recommendations, or if sales meetings don't reference metrics and trends, then literacy training will have little strategic effect. Start with teams that show interest, demonstrate value there and then expand more broadly.

Design for delivery and accountability

When designing a data literacy program, leadership should focus on three practical questions:

1. How will the organization deliver training?
2. How will it embed data literacy into workflows?
3. How will it measure progress?

Delivering data literacy should balance organizational programs and individual development. Structured pathways build a shared language and ensure broad coverage through workshops, curated learning paths or role-specific training modules. Individual development addresses depth and personal pace through online courses, conferences, mentoring and experiential learning. Effective programs use organizational delivery to establish a common foundation and individual pathways to develop the skills different roles require.

While delivery format matters, embedding training and practice in daily work has a greater effect. Programs that produce lasting change treat data as a natural part of work, not a separate training event. Ways to do this include:

• Designing meetings where relevant data is expected and discussed.
• Reviewing decisions with attention to what data was consulted and how it was used.
• Giving employees access to the data they need in formats they can use with tools they understand.

Many programs skip measuring progress or approach it casually. Establish a review cycle from the start to signal that the organization is learning alongside participants.

Before-and-after literacy assessments are useful but incomplete because they measure knowledge, not behavior. Better indicators of progress include changes in how employees conduct meetings, fewer requests for bespoke reports and reductions in data quality errors caused by users. Increasing adoption of analytics tools beyond spreadsheets is also a meaningful signal.

Programs grow with the organization

Data literacy programs don't have a completion date. Assuming they do can force repeat investment every few years when priorities shift, skills erode, the workforce changes or new technologies create literacy gaps. Organizations that maintain an ongoing approach to training will build teams that use data more effectively, question it more insightfully and communicate with it more widely.

Donald Farmer is a data strategist with 30+ years of experience, including as a product team leader at Microsoft and Qlik. He advises global clients on data, analytics, AI and innovation strategy, with expertise spanning from tech giants to startups. He lives in an experimental woodland home near Seattle.