Short answer
Begin with a process that is repeated often, follows clear rules and causes visible delays or errors. Map how it really works today, remove unnecessary steps, then decide what to automate, what to connect and what still needs a person.
Digital transformation can sound like a large programme. In practice, it usually starts with one process that everyone agrees is painful.
Choose the right first process
Good candidates share a few traits:
- Frequent: it happens daily or weekly, so improvements add up quickly.
- Rule-based: most decisions follow known rules rather than individual judgement.
- Visible: delays or errors are already noticed by customers, managers or staff.
- Contained: it involves a small number of teams and systems, so it can be improved without a long project.
Approvals, document handling, data entry from forms or emails, and recurring reports are common starting points.
Map how it really works
Ask the people who do the work to walk through it step by step. Note every handoff, every system, every spreadsheet and every place where someone waits for information.
The real process is often different from the documented one. Workarounds usually exist for a reason, and they show you where the current systems fall short.
Simplify before you digitise
Automating a poor process only makes it faster at being poor. Before building anything, ask which steps can be removed, combined or moved earlier. Duplicate approvals and re-checks of information that is already correct are common candidates.
Decide: automate, connect or redesign
- Automate repetitive, rule-based steps such as routing, notifications, validation and document generation.
- Connect systems where information is re-typed or exported between them.
- Redesign steps that exist only because of old constraints.
- Keep people where judgement, relationships or exceptions matter.
Roll out, then keep improving
Launch with the people who use the process, gather feedback in the first weeks and adjust. A modest improvement that the team actually adopts is worth more than an ambitious one that gets worked around.
Practical takeaway
- Pick one frequent, rule-based process with visible pain.
- Map the real process with the people who do the work.
- Remove steps before automating the rest.
- Plan for feedback and adjustment after launch.
