There's So Much You Can Do With AI — Where Do You Start?

DD Pim van den Boomen • 7 July 2026 •7 min

Pillar 03 // Where to start

start.md

$ start --where --not-what

There's so much you can do with AI, where do you start?

Start with the work that comes back most often and that only you do. Not with a tool, not with the smallest task, and not with whatever passes through your feed. First the place, then the tool. Most people do it the other way around and end up with a stack of subscriptions nobody looks at anymore. The question isn't what's possible, but what costs you the most.

DD DataDrift Digital • 7 July 2026 • 7 min

Overwhelm is the most common feeling in this market and the least discussed problem. There's something new every week, everyone's telling you that you're falling behind, and the advice you get everywhere is the same: start small.

That advice isn't wrong. It's just empty. It tells you how big your first step should be, not where you should take it. And the place determines whether after two months you have something that keeps running, or a trial subscription you forget to cancel.

01 / the definitionWhere do you start with AI in your business?

At the most expensive leak. That's the work that keeps coming back, requires little judgment, and yet is still done by you. Usually that's email, follow-up, or writing down what was discussed. You pick that place first, and only then the tool. Do it the other way around and you pick a tool and go looking for a problem to attach to it.

That's the whole order of operations. Everything below is an explanation of that one reversal.

02 / the ladderWhich rung are you on right now?

It helps to know where you're coming from, because most people are further along than they think and at the same time expect something from the next rung that it doesn't give. There are three.

01One-off questions. You paste something into a chat now and then and pull out a text. It works, it saves time, and nothing of it sticks around. Every time you start over explaining who you are.

02A chat that knows your business. You've written down your way of working once and use it over and over. The output gets noticeably better. But something still only happens when you open it and ask for it.

03A system with memory. Your way of working is stored in a place that multiple tasks draw from, and those tasks start without you. A draft comes in because there was an email, not because you opened something.

The difference between rung 2 and rung 3 isn't the intelligence of the model. It's who fires the starting gun. On rung 2 that's you, every time. On rung 3 that's an event in your business: a conversation that ends, an email that comes in, a lead that's been quiet too long.

The jump that matters isn't from dumb to smart. It's from "if I think of it" to "even if I don't think of it".

You can do rungs 1 and 2 yourself, today, for free. Do that too. Whoever buys rung 3 without ever having stood on rung 2 is buying something they can't judge the quality of.

03 / the adviceWhy is "start small" a half answer?

Because it's about size, not location. Start small on the wrong process and you have a small solution for a problem you didn't have. The advice comes from everywhere because it can never really be wrong, and that's exactly why it doesn't point you anywhere.

There's also a hidden assumption in it: that every process is equally worth tackling. That's not true. Automating five minutes a week is a hobby. Work that comes back every day and that you're doing in the evening is a business problem. Those aren't two sizes of the same thing, they're two different things.

What does point you somewhere is the question of which bottleneck is keeping your business small, and whether that bottleneck is even the kind a system can solve. That test is laid out in how do you know if you're ready for an AI system, including the cases where the answer is no.

04 / the choiceHow do you find the most expensive leak?

With three questions, and you don't need anyone else for it. Together they take half an hour and work better than any tool comparison you'll find online.

What did I do more than twice this week? Go through last week's sent mail and your calendar. Repetition is the raw material; without repetition there's nothing to hand off.
What structurally keeps getting pushed back? Not what's difficult, but what always gets postponed. That's almost never unimportant work, it's work without a deadline. Follow-up usually tops this list.
What can only I do? That's the part you specifically don't hand over. Everything outside of that is a candidate.

What's left after those three questions is usually one of three places: your inbox, your follow-up, or writing things down after a conversation. That's not a coincidence. Those are the three places in a small business where volume, repetition, and lack of a deadline come together.

We ourselves usually start with email, because within two weeks you can see whether it's right: you read each draft and know exactly what you would have written yourself. That reasoning is laid out in your inbox as your first AI employee. If your leak is in writing things down after conversations, then client files that keep themselves up to date is the better first step.

05 / the catchWhat goes wrong if you start with the tool?

You end up with a collection of separate things that each do something and together do nothing. Five subscriptions, three logins, none of them knowing what the others know. Each piece works on the day you set it up, and nobody notices when one of them stops working two months later, because there was never a moment when it had to happen.

The cost of that isn't the subscriptions. It's that after six months you think AI doesn't work for your business, while you never actually set up anything that could work. That's the most common way this goes wrong, and from the inside it looks like an honest experiment.

There's also an honest downside to rung 3, and we'd rather name it ourselves: every piece that runs needs maintenance. Something changes in your offering, an integration shifts, a supplier tweaks something. What runs needs to be kept up, by you or by someone else. Anyone who sells you a system without mentioning that word is selling you the first month, not the rest.

06 / actionableWhat do you do this week, without buying anything?

Three things, in this order. Together they cost a few hours and they produce something that lasts, even if you never buy anything.

01Point to the leak. The three questions above, with your calendar and sent mail next to you. Write down one sentence: this is the work that comes back most often and costs the most.

02Describe that one process. Do it consciously three times and note what you did, in what order, and why you chose what you chose at every fork. That why is the part that usually never comes out.

03Test it by hand. Use that description for a week in an ordinary chat. If it doesn't work with your own description right there, a more expensive version won't save it either.

What you're left with is a description of one process in your own words. That's the building block for everything that comes after, and it's also the only thing a provider can actually build on. Whether you then come to us or to someone else, the conversation starts two steps further along.

Frequently asked questions

What's the difference between ChatGPT and a custom AI system?+

The difference isn't in the model but in the source and the trigger. A chat knows what you tell it in that conversation and then starts blank again. A system draws from your recorded way of working and starts on its own, because an email came in or a conversation ended. Same technology, different place in your business.

Which process should I automate first?+

The process that comes back often, requires little judgment, and is still done by you. In small businesses that's almost always email, follow-up, or writing things down after a conversation. Don't pick the process that sounds most exciting, and not the smallest one either, but the process that happens most often.

Can I start with free tools?+

Yes, and for the first two rungs that's actually the smartest move. A free chat plus your own description of your way of working gets you further than most paid tools. Only once you notice that the bottleneck is that something only happens when you open it yourself, is there a reason to look further.

How long before something is actually running?+

That depends on the process, and nobody can tell you in advance without having seen your work. What is predictable: recording your way of working takes a block of time upfront, followed by weeks of corrections. A provider who names a number of weeks before knowing what you do is naming an average that isn't about you.

Should I wait until the technology gets better?+

No, because the part that's changing isn't the part you're working on. The models keep getting better and that will continue. What doesn't change is that you need to know which work can be recorded and which work stays with you. Figuring that out takes time and doesn't go stale, even if you only build something a year from now.

30 minutes, free, no sales pitch

Not sure which leak is the most expensive one for you?

Schedule a call

We'll look at where your time is leaking and tell you honestly whether we can do something about it. Often the answer is: you can do this yourself. Then we'll say so, and you'll have spent thirty minutes on a clear answer.

→ Schedule a call (30 min)

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