This is the post in which we explain when you shouldn't call us. Not out of modesty, but because the wrong client at the wrong time is an expensive mistake for both parties.

The question usually arrives less neatly and sounds like: everyone's doing something with AI and I'm not yet, am I too late? That's a feeling, not a diagnosis. Below is the diagnosis: which bottleneck you need to have before a system means anything for you, and which bottlenecks it actually makes worse.

01 / the definitionWhat does "being ready" mean for an AI system?

Being ready means you have a way of working that repeats. You sell something to someone you can name, you deliver it in a way you've already done twice before, and it costs you time you no longer want to give. A system captures that way of working and runs work through it. It doesn't invent the way of working.

That's the whole test. Not your turnover, not your team size, not whether you're good with computers. The question is whether there's something that goes the same way often enough to hand over.

02 / the testWhich bottleneck is keeping your business small right now?

Almost every business that isn't growing gets stuck at one of five points. Which of the five is at the top for you determines whether a system solves something or only speeds something up. Go through them and be strict: there's always one that weighs heavier than the rest.

A system strengthens a way of working that's already running. It doesn't invent one.

03 / the boundaryWhen is it too early?

It's too early as long as you can't say in one sentence what you sell to whom. That sounds harsh, but it's a predictable failure in this market. Without a fixed offer there's no way of working to document, only a collection of loose attempts that come out just as loose in a system.

This isn't our idea. It's what everyone in this category says as soon as the conversation becomes honest, including people who have an interest in selling you something. That doesn't make it a marketing line but an observation: it goes wrong in the same place.

Three situations in which you'd better wait:

• You're still working out what your offer should be. Wait until you've sold it three times to three similar people.
• You have too few clients and that's the problem. A system doesn't turn too few clients into more clients. Fix the gap at the front end first.
• You've just bought something expensive that delivered nothing and you're still recovering. Then buy nothing, not even this. Go and find out first why the last thing didn't land, because that reason doesn't disappear through another purchase.

You won't hear that last point from a supplier any time soon. We do hear it in conversations, literally: if I add up what it's cost so far, the question is whether I'm even going to get that back out of it. Someone in that position doesn't need a new system. They need peace and one thing that demonstrably works.

04 / the catchWhat does it cost you even if you are ready?

Two things you won't easily find in a brochure. The first is a block of time upfront: you have to explain how you work, and that's work. The second is a fixed moment each week where you feed in what's changed about your offer, your target audience or your way of doing things. Without that moment, it quietly goes stale.

Then there's the correction period. In the first weeks, the drafts are half right. They're too formal or too loose, they're missing context that only you have, or they neatly answer the wrong question. Reviewing them can sometimes take just as long as doing it yourself. That's not a malfunction, that's the period in which your judgement enters the system.

Anyone who promises it'll be right from day one has never done it. Anyone who promises it'll run itself after that, hasn't either.

05 / the signHow do you know you actually are ready?

By irritation, usually. You notice you've given the same explanation for the third time this week. You notice you have to personally touch every assignment before anything happens. You notice the work that keeps piling up is always the same work. Those aren't problems of too few clients. Those are problems of too much of the same.

There's a second sign, and it's sharper: you can explain what you do to someone who has to take it over. Not in a manual, but out loud, in an hour. If you can do that, there's something to document. If you can't, because it's completely different for every client, it's worth first checking whether that's really true. Usually there turns out to be more pattern to it than it feels like from the inside.

06 / usableWhat do you do if you're not ready yet?

Something small and dull, and you don't need anyone for it. Pick one thing you do often that irritates you. Do it three times deliberately and write down what you did, in what order, and why you chose what you chose at each turn. That last part is the bit that normally doesn't come out, and that's exactly the bit you need.

What you're left with is a description of one process in your own words. That's the building block. Whether you later have us use it, or someone else, or do it yourself, doesn't matter: without that description, no supplier has anything to work with, and with that description, you immediately have a much sharper conversation.

If your offer isn't fixed yet, that's your first project. Not with us, and not with a system. Just by selling it until it becomes sharp.

This text was produced with AI assistance and checked and approved by a human before publication.