If an audit starts from the tool, it often produces a polished automation promise. If it starts from the actual workflow, it becomes clear whether the first need is automation, simplification, better data, or clearer ownership.

1. Describe real work first

Map what people do every day: where information comes from, where it is retyped, who approves it, and where it waits.

2. Separate error causes

Errors can come from manual entry, weak data, unclear responsibility, or poor system logic. Each cause needs a different fix.

3. Follow the data journey

If information moves through Excel, email, ERP, CRM, and warehouse tools, the audit should show where it is duplicated or loses trust.

4. Count more than hours

Saved time matters, but so do error cost, delays, decision speed, customer service, and employee load.

Practical principle: automate only a process whose rules can be explained simply.