An activity-led future-of-work strategy begins with a deceptively simple question: what dopeople actually do?

Job descriptions are useful, but they are not enough. They typically describe purpose, accountbilities, and broad responsibilities. A responsibility such as “provide strategic advice” may contain research, data gathering, interpretation,stakeholder interviews, option generation, drafting, challenge, and follow-up.Those are very different activities with different AI potential.

A practical activity map does not attempt to document every mouse click. It creates a structured, reusable representation of the meaningful work that makes up a role. For most roles, an initial map of 10 to 20 activities is sufficient to begin a high-quality discussion.

Start with four inputs.

First, the role itself. Review the job description, performance objectives, outputs, and decision rights. Identify the responsibilities that define the role’s value.

Second, the process context. Existing process maps, operating procedures, and systemdocumentation show where the role participates in recurring workflows. Use themas evidence, not as the complete answer.

Third, employee and manager insight. Short, structured conversations reveal the gap between formal design and lived work.

Ask: What do you spend time doing? Which activities repeat? Where does work get stuck? What requires judgement? What is invisible in the official description?

Fourth, operational data where appropriate. Case volumes, ticket categories, system logs, calendar patterns, and work samples can validate the map. Use these responsibly: the purpose is understanding work, not covert surveillance.

Each activity should be written in a consistent verb-object form:

1. “Review contract clauses”,
2. “Prepare management report”,
3. “Resolve customer exception”,
4. “Coach team member”.

The wording should be concrete enough for an employee to recognise and broad enough to recur across roles.

Then tag each activity with a small number of useful attributes:

• Time share or relative importance.
• Work type: transactional, cognitive, relational, or physical.
• Degree ofstructure and repeatability.
• Data, systems, or documents used.
• Judgment, risk, and accountability required.
• Skills drawnupon.

The result is more than a list of tasks. It is a model of the role’s work content. It allows leaders to see where time is consumed, where frustration sits, what work is central to value, and where technology could alter the activity mix.

Keep the map alive.

Work changes as policies, systems, customer expectations, and tools change. An activity map should be versioned and reviewed, not treated as a one-time job-analysis exercise. A credible programme starts small: select a role family, build a first-cut map, validate it with practitioners, and improve it through use.

The objective is not perfect granularity. It is a common, evidence-based language for discussing work before deciding how to change it.

**Next article:**External occupational taxonomies can accelerate this work—provided they aretreated as scaffolding rather than a substitute for internal reality.