Which business process should you automate first?
The best first automation is usually a task the business understands well enough to describe, test and supervise. Look for repeated work with a worthwhile cost, clear information and manageable exceptions. Start by deciding what should happen before choosing a tool to make it happen faster.
Editorial responsibility: Nardus Oelofse, Chief Consultant.
By Chase & Marshal
Look for repeated effort with a clear outcome
Ask teams where they copy information, chase routine updates or move a request between systems. Record how often the task occurs, how long the complete workflow takes and what errors or delays matter. The objective is to identify a meaningful burden.
Separate the task from the surrounding process. A small automated action can be useful, but it may have little impact if the main delay occurs elsewhere. Follow the work through to the customer or colleague receiving the result before estimating the benefit.
Check whether the rules are stable
Describe the normal route and the conditions that change it. Ask an experienced staff member what they consider when handling an exception. If a task depends on judgement that nobody has made explicit, further process work may be needed first.
In an illustrative finance administration task, routing a complete request to the correct approver may be straightforward. Deciding whether an unusual expense is justified requires a different level of authority. Automate the reliable movement of work while keeping the consequential decision clear.
Include the cost of maintaining the change
Assess the quality of the source information and the reliability of the systems involved. Consider what happens when a field changes, a connection fails or a person leaves. Someone needs to monitor exceptions and maintain the workflow.
Compare the likely benefit with setup, testing, training and ongoing support. Time saved is useful when people can redirect it to worthwhile work or avoid a real capacity constraint. Be precise about that benefit rather than treating every automated minute as an immediate cash saving.
Run a bounded test with a fallback
Choose a representative set of cases and define what an acceptable result looks like. Include missing data, duplicates and other foreseeable exceptions. Agree who reviews the output, how errors are corrected and when to pause the automation.
Measure the full process after the change, including supervision and rework. If the result is useful, document the operating responsibilities before expanding. A good first automation builds both a practical improvement and the organisational capability to make the next decision more confidently.
Questions answered
Should we automate the task taking the most time?
Consider it, but also assess the rules, information, exceptions and consequence of error. A smaller, well-understood task can be a better first project if it creates worthwhile value and is easier to maintain.
Related
Sources
Sources supporting the factual statements in this article. Findings should be read within their stated dates, geography and scope.
- Helping Kiwi businesses succeed - Business.govt.nz. www.business.govt.nz. Publication date not supplied.
Choose a task worth automating
Bring one repeated task that consumes time or creates avoidable errors. We can help assess its suitability and define a practical first automation.
