"Where should we start?" is the question we hear most often from businesses considering their first automation pilot. Here is a practical framework for answering it.
What makes a good automation candidate
Not every workflow is a good fit for automation. The best candidates share four characteristics:
- High frequency. It happens often enough that the time savings add up. A daily task saves more than a quarterly one.
- Rule-based. The steps follow a predictable pattern. If-then logic, not judgement calls.
- Time-consuming. Someone is spending meaningful hours on it each week or month.
- Clear output. The result is defined and measurable. A report, a notification, a document, a summary.
The scoring table
List your candidate workflows and score each one on three dimensions, from 1 (low) to 5 (high):
- Frequency: How often does it happen? Daily = 5. Monthly = 2. Quarterly = 1.
- Friction: How painful is it to do manually? Hours of assembly work = 5. Quick copy-paste = 1.
- Output clarity: How well-defined is the result? Standard report format = 5. "It depends" = 1.
Multiply the three scores. The workflow with the highest total is your best starting point.
Common examples
These are workflows that score well across most operational businesses:
- Weekly compliance reports. High frequency, predictable format, clear output. Often the single best first project.
- Appointment follow-up sequences. After a meeting or consultation, a predictable set of messages need to go out. High frequency, rule-based, clear output.
- Shift summaries. End-of-shift handover documents that pull from existing logs and systems.
- Quote chasing. Following up on sent proposals at defined intervals. Entirely rule-based.
What does not score well
Workflows that require heavy judgement, have unpredictable inputs, or produce different outputs every time are poor automation candidates. Sales conversations, complex negotiations, and creative work are examples. These stay human.
Your next step
Take 20 minutes to list your top five repetitive workflows. Score each on frequency, friction, and output clarity. The math will point you to the right starting place.
