What would automating this give back?
Four numbers and six questions. You get the hours a year, the money, and the share of it automation realistically returns for a task like yours.
THE TASK
How often does it run?
How much of it is a rule rather than a judgment?
When did the process last change?
How many people does it pass through?
What does a mistake cost?
Is the work described anywhere?
WHAT IT GIVES BACK
Answer the six
The cost of the task is the easy half. The share of it you get back is the half that decides anything.
A SECOND OPINION
Not sure it is the right task?
Send the number with two facts about your business and we reply within one business day with what it usually means at that size.
Most calculators count the whole task as saved. Automation never returns the whole task. This scores how often it runs, how much is rule versus judgment, and how settled it is, then counts only that share.
What decides the answer
Frequency
A task that runs fifty times a week returns more than the same task run twice, because the saving is per run and the build is once.
Rules against judgment
Work that follows a rule automates. Work that needs somebody to decide does not, and pretending otherwise is how automation projects get abandoned.
How settled it is
A process that changed last month will change again. Automating an unsettled process means maintaining two of them.
Handoffs
Every handoff is a place work waits. The waiting is usually the larger half of the cost and almost nobody counts it.
Cost of a mistake
Where being confidently wrong is expensive, a person stays in the loop and the saving is smaller. That is a real constraint, not a failure.
How many hands
A task that passes through four people has four places it can wait. Those are the hours nobody counts, and they are usually the larger half.
Why the honest number is smaller
Ask any automation calculator what a task is worth and it will multiply minutes by frequency by an hourly rate and hand back the total. That number is the cost of the task. It is not the saving, and the gap between the two is where automation projects go to die.
Automation returns a share of a task, never all of it. Somebody still handles the exceptions. Somebody still checks the run that looks wrong. Something still breaks when the supplier changes a form. The share you get back is decided by the shape of the work rather than by the tool, and the shape is knowable before anything is built.
So this asks about the shape first. A task that runs constantly, follows a rule, has not changed in a year and passes through three people gives back most of its hours. A task that runs twice a month, needs a judgment call, changed last quarter and is expensive to get wrong gives back very little, and the useful answer there is to leave it alone and fix something else.
The arithmetic is printed under the answer. Every assumption in it is on screen and editable, because a number you cannot check is worth less than no number.
Questions we get
How accurate is this?
It is an estimate built from four figures you supply and six answers about how the work runs, and it shows every step of its arithmetic. It is accurate enough to decide what to look at first and not accurate enough to put in a budget. The way to improve it is to log one week of the task rather than to estimate harder.
Why is the number lower than other calculators give?
Because the others report the cost of the task and call it the saving. This returns only the share that a task of that shape gives back, which is smaller and closer to what happens.
What if the task scores badly?
That is an answer too, and a useful one. A low score usually means the process is not settled. Fix it by hand, write down what you did, and score it again in a month. Automating an unsettled process is how you end up maintaining two of them.
Does anything get sent anywhere?
No. It runs in your browser, nothing is transmitted, and there is no sign up. You can close the tab and the answer goes with it.
What do I do with the number?
Rank your candidates with it. Running one task through tells you whether it is worth doing. Running five tells you the order, which is the more useful answer.
More in the guides and every answer in one place.
Shaheer leads the work, with engineers, writers, filers and analysts behind him. C-suite operations for a San Francisco AI company, Six Sigma on the process side, Anthropic certified on the Model Context Protocol, ten years across eight industries. See what we have built