Measure the After and Time Saved
Compare your before and after numbers to show honestly what your automation changed.
Objective
By the end of this lesson, you can calculate weekly hours saved from your baseline and after measurements, and report the result honestly.
Measure the same way twice
In Module 2, you timed the workflow before any change. Now you time it again, the same way, after the automation has run for at least one full week.
Use the same measures:
- Minutes per item, including the human steps that remain
- Items per week
- Response time
- Error or miss rate
Do not count only the automated part. If a person still spends two minutes approving each draft, that time counts.
Time is not the only result
Report what changed, even when it is not time:
- Response time. Did families hear back in one hour instead of one day?
- Misses. Did zero requests fall through the cracks this week?
- Errors. Fewer wrong phone numbers?
- How it feels. Ask the worker. "Do you dread this task less?" Their words count as evidence.
A local example
The taquería's catering intake, before and after:
| Measure | Before | After |
|---|---|---|
| Minutes per request | 15 | 4 (reading the summary and approving the draft) |
| Requests per week | 12 | 12 |
| Hours per week | 3.0 | 0.8 |
| Average response time | About 20 hours | About 2 hours |
| Missed requests | 2 of 12 | 0 of 12 |
The owner's daughter put it this way: "I still answer every family. I just don't copy anything anymore."
Try the time-saved calculator
Before and after
Put in your baseline and after numbers. Run it to see weekly and yearly hours saved. Keyboard users: press Esc, then Tab, to leave the editor.
Notice the upkeep line. Automations need care. Count that time, or your result is not honest.
Report it honestly
- Say how many runs you measured and over how many days.
- Say what still needs a person.
- If the result is small, say so. A small, true result beats a large guess.
- If it made something worse, say that too, and what you changed.
Reflection
Go deeper: AI, Minus the Hype by Dom Davis, Temerarii Media. Its section on measuring AI like any other project fits here.