What Should You Automate First?

Start with the process, not the technology

When businesses explore automation, the first question is often, “What can we automate?” A better question is, “Which process is worth improving?”

Repetition alone does not make a task a good candidate. The strongest opportunities combine meaningful manual effort with business impact, a stable process, and realistic feasibility.

What makes a good automation candidate?

• It happens frequently enough for the improvement to matter.
• It consumes meaningful employee time.
• Delays or errors create rework, cost, or missed opportunities.
• The process follows reasonably consistent steps.
• The rules and inputs can be understood clearly.
• The systems involved can be connected reliably.

Evaluate the opportunity across seven dimensions

These are evaluation dimensions, not steps in a sequence. Use them together to compare potential candidates and make the trade-offs visible.

• Frequency — How often does the process happen?
• Effort — How much human time does it consume?
• Impact — What business outcome does it affect?
• Stability — Does the process work in a reasonably consistent way?
• Rule clarity — Can the decisions and steps be defined?
• Feasibility — Can the required systems and data be connected?
• Exception rate — How often does the process require handling outside the normal rules?

Value matters, but so does implementation risk

The highest-value process is not always the best first automation. A process may have significant savings potential but require complex integrations, frequent exceptions, or substantial process redesign.

For a first project, the better choice is often the process with the strongest combination of business value, feasibility, and low implementation risk. A slightly smaller opportunity that can be delivered and measured quickly may be more valuable than a much larger project that remains difficult to implement.

How to score the candidates

For a simple comparison, score each dimension from 1 to 5: 1 means weak, 3 means moderate, and 5 means strong. For Exception rate, reverse the scale: 5 means very few exceptions and 1 means frequent exceptions.

The score is a comparison tool, not a pass/fail test. Look at the individual scores as well as the overall picture; a high score should not hide a major feasibility, risk, or exception problem.

A worked example: lead follow-up

Consider a business where every new enquiry is read by an employee, copied into a spreadsheet, checked against previous interactions, answered, assigned, and given a follow-up reminder.

This is a strong candidate because it happens repeatedly, involves information moving between systems, and delays can result in missed follow-ups. The routine steps can be automated while judgment-heavy conversations remain with people.

A useful baseline could include response time, follow-up completion, manual time per enquiry, and the number of missed or delayed enquiries.

Fix the process before you automate it

Automation can make an inefficient process run faster without making it better. First understand the current workflow and remove unnecessary approvals, duplicate entry, and avoidable handoffs.

A practical sequence is: Understand → Simplify → Standardize → Automate.

When should you not automate?

• The process happens too rarely to justify the effort.
• Every case is substantially different and requires judgment.
• The process itself is poorly defined.
• The cost or risk of an automated error is too high.
• The manual process is already cheaper and sufficiently effective.

Where should you start?

List five to ten manual processes and score them against the seven dimensions. Then investigate the strongest candidates, paying particular attention to business value, feasibility, and implementation risk.

Start with the automation you can prove, not simply the one that looks biggest. A recurring report, data transfer, approval, invoice workflow, or follow-up process can be a better first project than a larger but more complex opportunity.

The takeaway

Choose a process where the business problem is clear, the workflow is stable, and the improvement can be measured. Establish the baseline, simplify the process, automate the repeatable work, and measure the result.

A successful first automation does more than save time. It gives the business evidence that automation works—and a clearer basis for deciding what to automate next.