The four hours a week your company is losing and has stopped noticing
A friend of mine runs a company with fourteen people. Good business, profitable, been going nine years. I asked him once where his week goes and he gave me the answer everyone gives, which is that it goes everywhere and nowhere.
So I asked a narrower question. What did you do on Monday morning?
He pulled figures out of three systems, put them in a spreadsheet, compared them to last week, wrote a short summary, and sent it to four people. Took him about ninety minutes. He has done that every Monday for six years.
That is roughly one hundred and thirty working days of his life, spent moving numbers from one screen to another.
He was not annoyed about it. That is the part worth paying attention to. He had stopped seeing it entirely, the way you stop hearing a fridge.
Why you cannot see it
The work that is worth automating is almost never the work that irritates you. Irritating work gets fixed, because irritation is a signal and people act on signals. What survives is the work that has quietly become part of the furniture.
Three things make it invisible.
It is distributed. Ninety minutes here, twenty minutes there, an hour on Thursday. No single instance is big enough to justify a project, so no single instance ever gets one.
It is nobody’s problem. The person doing it usually accepts it as the job. The person who could authorise fixing it does not see it happen. The two rarely have the conversation, because from the outside it looks like the process is working. And it is. Working is not the same as worth doing.
And it feels like the price of doing business. Someone has to reconcile the invoices. Someone has to answer the same email. The assumption is not that this work is unavoidable, it is that nobody has ever questioned whether it is, which is a different thing and much easier to fix.
The method
Here is what I do at the start of an engagement, and there is nothing in it you cannot do yourself on a Tuesday afternoon.
Do not ask people what takes them the longest. They will tell you what frustrates them, which is not the same list. Frustration correlates with difficulty, not with hours. The thing that eats the most time is usually something they find quite easy and mildly pleasant, because easy repetitive work is restful and people do not resent it.
Ask a different question: what do you do every week that you could describe to a new hire in under five minutes?
That question is doing something specific. If a task can be fully explained in five minutes, it has few branches, few judgement calls and a predictable shape. That is precisely the profile of work a machine handles well. Anything that takes forty minutes to explain has exceptions buried in it, and exceptions are where automation projects go to die.
Then ask when they last had to think hard while doing it. If the honest answer is “not for a couple of years”, it has stopped being work and started being a ritual.
Then ask what happens to the output. Sometimes the answer is that nobody reads it. I have found reports being produced weekly for people who left the company. Not often, but often enough that it is worth asking.
Scoring what you find
You will end up with a list of somewhere between ten and thirty things. Most of them are not worth touching. Three numbers sort them.
How often does it happen. Daily beats weekly beats monthly, by a wide margin. Something that happens twice a year is almost never worth automating no matter how tedious it is.
How long does it take, honestly, including the part where you find the password, the part where you wait for the export, and the part where you get distracted afterwards and take six minutes to get back into what you were doing. People underestimate this by about half.
How mechanical is it. Score it one to five, where five means a competent stranger with your instructions would produce exactly the same output as you, every time.
Multiply the first two, filter by the third. The top of that list is your answer, and it is very often not what anyone in the room expected.
What does not belong on the list
I will save you some disappointment.
Anything where the input arrives in a different shape every time. Automation handles variety badly and handles chaos not at all. If forty per cent of your supplier invoices arrive as photographs of paper, that problem is upstream and no amount of clever tooling fixes it from this end.
Anything where being wrong is expensive and being right requires context nobody has written down. Pricing decisions, hiring calls, anything that touches a relationship with a customer who thinks they have a relationship with a person. These can often be assisted, where a system prepares the decision and a human makes it, but they should not be handed over.
Anything that only one person understands and that person is leaving. Fix the documentation first. Automating an undocumented process means encoding whatever that one person happens to do, including the mistakes.
And anything you are doing because of a regulation you have not read since 2019. Check first. Some of it may not need doing at all, which is cheaper than any automation.
What to do with the list
Take the top item and fix that one. Only that one.
This is the part most companies get wrong. Having done the exercise and produced a satisfying list of twenty things, the instinct is to plan a programme. Do not. Fix the single highest scoring thing, live with it for a month, and see whether it actually held. Automations fail in ways that only show up when a real edge case arrives, and the first one teaches you more about your own processes than the list did.
Then do the next one.
My friend’s Monday report now builds itself. It took an afternoon. He still reads it and still writes the last two lines himself, because that part is judgement and judgement is his job. He got back about seventy minutes a week, which is not a dramatic number, and he says it is the single most noticeable change to his working life in three years.
That is what this work usually looks like. Not transformation. Just the steady removal of things that were never worth a person’s attention.
If you want someone to go through this with you, that is more or less what the first call is. Thirty minutes, and you can take the list and do it yourself. Book a call.