Free · nothing to sign up to

What is it actually worth?

Ten minutes. At the end you will know the one job in your business worth handing over first, roughly what it is worth to you in a year — and, just as usefully, whether the answer is “not yet”. Nothing is sent anywhere; the arithmetic happens in this page.

How the sums work

No black box, and nothing flattering hidden in it. Three lines:

hours a year = times per week × minutes each time × 46 ÷ 60 what it costs you = hours a year × your hourly number what it is worth = that, halved

Forty-six weeks, not fifty-two, because you do take some time off, even if it does not feel like it.

And then it is halved. That is the part a vendor would quietly leave out. You will not get all the time back: you still have to read the output, fix the bits it got wrong, and occasionally start again. Halving it is the honest version — and it is still usually a number that makes the decision for you. The halved figure is the one shown above in large type; the un-halved one is the footnote.

Why the four questions decide it, not the money

The job with the biggest number is very often the wrong place to start. What matters is whether the job is ready, and that is what the score out of twenty measures.

  • Repetition. The whole return comes from repetition. A job you do twice a year is not worth automating however annoying it is.
  • Sameness. The one people get wrong. Quoting scores high because the structure repeats even though the prices do not. Giving a customer advice scores low, because the thinking is different every time.
  • Whether a draft is fine. Good at first drafts, bad at final answers. A job you were always going to check is a safe job to hand over. A job where nobody checks is where the trouble starts.
  • Whether it needs to know things only you know. This is the one that kills most enthusiastic ideas. It does not know your business, your history or your customers unless you tell it — every single time.

Score under twelve and the honest answer is not this job, not yet. That is what the tool will tell you, on a page that would plainly rather sell you something. A low score means the job is not ready, not that you are behind.

A worked example

A letting agent writes six tenancy summaries a week, twenty-five minutes each. Her hourly number is £55.

6 × 25 × 46 ÷ 60 = 115 hours a year. At £55 that is £6,325. Halved, about £3,100 a year — from one job.

She scored it 5 for repetition, 5 for sameness, 4 for draft tolerance and 4 for self-contained: eighteen out of twenty. It was the first thing she handed over, and it took an afternoon to set up.

Questions about the sums

Why forty-six weeks and not fifty-two?
Because you do take some time off, even if it does not feel like it. Fifty-two weeks would quietly inflate every figure on this page by about thirteen per cent.
Why halve the number?
Because you will not get all the time back. You still have to read the output, fix the bits it got wrong, and occasionally start again. The halved figure is the honest one, and it is the one you can defend to whoever signs for it.
What if my best job scores under twelve?
Then do not start, and do not start anywhere else either. A low score means the job is not ready, not that you are behind. Usually it means the job depends on things only you know, which is fixable, but by writing things down rather than by buying software.
Do I have to give you my email to see the result?
No. There is no form, no gate and no download. The answer appears on the page as you fill it in.
Where do my figures go?
Nowhere. The arithmetic runs in your own browser and nothing you type is sent to us or to anyone else. Close the tab and it is gone.
Is this worth doing if I only want to automate one small thing?
That is exactly what it is for. One job, done properly, is the only approach that reliably works; the people who get nothing out of this are the ones who tried six at once.

The rule that saves you a fortnight

Do one job. The highest scorer. Only that one, for a fortnight.

Everyone who gets value out of this does one thing properly before adding a second. Everyone who gets nothing out of it tried to do six at once in a burst of enthusiasm in week one, got mediocre results at all six, and concluded the technology does not work.

On day one, take the last five examples of that job that you did yourself — real ones, with the real messy input and the finished version you actually sent. Show it those five. Not a description of what you want: the actual before and after. Then give it a sixth input and compare what comes back against what you would have written. That is the whole test.

If the number made the decision for you

Ask me about that job directly — tell me what it is and I will tell you honestly whether it is the sort of thing I am good at, and what I would need from you first.

Try me