Data literacy by role: what each profile in your company needs to know
Not everyone needs to know the same things about data. An executive doesn’t need to code, and an analyst doesn’t set the strategy. Defining what each role should know is what makes training useful rather than a box-ticking exercise.
Why by role
A role-based framework answers a simple question: what does this person need to be able to do with data to do their job well? That keeps training focused on what each person uses, and gives measurement a clear reference point.
Four levels
1. Basic: all staff
- Understand what a data point, an indicator and personal data are.
- Spot a misleading chart, such as an axis that doesn’t start at zero.
- Know what data can be shared and with whom, and what data must not go into an AI tool.
2. Use: managers and heads of area
- Read a dashboard and know what to ask before deciding.
- Tell correlation from causation, and a trend from a one-off.
- Ask for data with a clear definition: what is counted, from when and from which source.
3. Analysis: analysts and business profiles who work with data
- Prepare and clean data, and document what they did.
- Check data quality before using it: gaps, duplicates, odd values.
- Explain a conclusion with its limits: what the data shows and what it doesn’t.
4. Management: data owners, data stewards and technical teams
- Define quality and access rules for the data in their domain.
- Maintain the glossary, catalogue and lineage of their data.
- Understand what regulation requires of their data (GDPR, and Article 10 of the AI Act if it is used to train AI).
Examples by role
- Executives: level 2, with emphasis on asking good questions before deciding.
- Sales or customer service: level 1, with emphasis on personal data and AI use.
- Finance or controlling: level 3, because a misreading costs money.
- Data owner or data steward: level 4 in their domain.
How to use the framework
- Assign a target level to each role in your organisation chart.
- Measure the current level with a short survey per role.
- Train first where the gap between current and target level is largest and most costly.
- Review the framework once a year, or when new tools arrive, such as generative AI.
Frequently asked questions
How many levels should we have?
Three or four. With more, the difference between levels becomes unclear and the framework is hard to maintain.
Are data literacy and AI literacy the same thing?
No, but they go together. Data literacy is working with data with sound judgement. AI literacy is using AI with sound judgement and knowing its risks. The second builds on the first.
Informational content.
What level of data literacy does your team have?
Free 5-minute assessment: your score on the 4 dimensions and what to work on first.