Before you automate anything, answer this: what would you do with the time?

Every conversation about AI at the moment circles the same worry: will it replace me? It’s an understandable question, and I think it’s the wrong one to spend your energy on. There’s a far more useful question sitting right next to it, and almost nobody asks it out loud:

If something gave you back six hours a week, what exactly would you do with them?

Try answering it properly. Not “get on top of things” or “be more productive” — those aren’t answers, they’re moods. I mean the specific thing you’d do on Tuesday afternoon once the admin stopped eating it.

Most people can’t answer immediately. That’s not a character flaw; it’s just that nobody plans for time they don’t have. But it matters more than which tool you buy, because of an unglamorous truth about freed-up time: it doesn’t stay free.

Unclaimed time grows admin back

Take four hours of invoice chasing off someone’s week and one of two things happens. Either those hours go somewhere they’d chosen in advance — or the week quietly closes over the gap. A few more emails get written at greater length. A report nobody reads gets prettier. A meeting expands to fill the morning now available to it.

Then, six months later, the honest verdict is “we tried automating some things, I’m not sure it made much difference.” And they’re not wrong. The hours really were removed. They just weren’t spent.

This is why I think the question comes first, before the tool, before the plan, before anyone builds anything.

The three answers that actually hold up

In practice, nearly every good answer I’ve seen is a version of one of three things.

• More of the work you want. The freed hours go into winning and doing better-fitting jobs — answering enquiries while they’re warm, following up properly, quoting the same day instead of next week.
• Better work, not more of it. The hours go back into the craft: more care per job, fewer rushed corners, the finish you’d like your name on. Same volume, higher standard, better reputation.
• A shorter week. The hours go to you. You stop doing paperwork at nine at night. This is a completely legitimate answer and I wish more people said it out loud.

Notice they lead to different decisions. “More of the work I want” means point the agent at enquiries and follow-ups. “Better work” means clear the admin sitting either side of the skilled part. “A shorter week” means start with whatever specifically keeps you up late. The same capability, aimed at three different targets, produces three different businesses.

Why this is the difference between a tool and a change

Automation is only ever a transfer: work moves from a person to a machine, and something has to arrive in the space left behind. If you’ve decided what that something is, you get a genuine change in how the business runs. If you haven’t, you get a slightly tidier version of the same week.

It’s also the honest reason I’m wary of the “replace people” framing. The interesting move was never removing the person — it’s freeing them to do the part a machine can’t. But that only works if you know which part that is, and you’ve protected the time for it.

Answer it in one sentence, then start small

You don’t need a strategy document. One sentence will do: “If I get Thursday mornings back, I’ll use them to ring the customers who haven’t heard from us in a year.” That sentence tells you what to automate, in what order, and how you’ll know whether it worked — because you can check whether the calls actually happened.

Then hand over one small job and see, which is a fortnight’s work rather than a leap, and treat the freed slot like a real appointment. Defend it as you would a meeting with your best customer, because in effect that’s what it is. For what it’s worth, my own answer is that the time I save the people I work with goes into the thinking I can’t do for them — and I’d rather they spent it well than let me quietly hand it back as busywork.