The 30-45 Rule: How to Find Your First AI Win

Find a task that takes you 30 to 45 minutes every day, or most days, that you do not enjoy and that follows a pattern. That is where AI should start. That is the whole rule. Not the most exciting task, not the most impressive one, not the project that sounds good in a board meeting. The boring one you do over and over that you would happily never do again. Point AI at that first, and you have found your first win.

Most owners get stuck on where to begin with AI because they reach for the big, ambitious thing. The 30-45 Rule does the opposite. It tells you to start with something small, frequent and dull, because that is the safest and fastest way to prove AI works in your business. Here is why that sweet spot is the right one, and how to spot yours.

Why 30 to 45 minutes is the sweet spot

The number is not random. It sits right where three things line up.

It is frequent enough to pay back. A task you do every day adds up. Thirty minutes a day is two and a half hours a week, which is over a hundred hours a year from one person on one task. Automate that and the return is obvious and quick. A task you only do twice a year is not worth the effort of handing it over, no matter how annoying it is.

It is small enough to trust a junior with. AI is like a junior employee you just hired. Clever, tireless, eager to help, but you would not hand a new hire the whole business on day one. A 30 to 45 minute task is low stakes. If it gets something wrong early on, you catch it, you correct it, and nothing breaks. That low risk is what lets you build trust before you expand.

It is repetitive enough to automate. Tasks in this band tend to follow the same steps every time. Same inputs, same logic, same output. That pattern is exactly what AI handles well. The task that changes completely every time is hard to hand off. The one that runs the same way each day is ready for it.

Anything much smaller than this band is not worth the effort. Anything much bigger usually carries too much judgement to trust a new system with straight away. The 30 to 45 minute daily task is the safe middle.

What this looks like across different businesses

The rule is industry-agnostic. Every business that sells the time of qualified people has a few of these tasks hiding in the day. Here are the ones we see most:

• Email triage. Sorting the inbox each morning, working out what each email is about, and filing it where it belongs. It is repetitive, it follows a pattern, and nobody enjoys it.
• File notes and write-ups. Turning a meeting or a consultation into a clean written record. Same structure every time, just different details. A perfect pattern for AI to draft and a person to check.
• Quotes. Pulling together a quote from a standard template, with the right line items and the right numbers. Done dozens of times a week in most service firms, and almost always the same shape.
• Report drafting. The first pass of a recurring report. AI builds the draft from the data, a person reviews and signs off. The blank page is the slow part, and that is the part AI removes.

The common thread is that the task runs the same way each time and a human still checks the output before it goes out. That is the shape you are looking for.

A 30-45 task that became $1.3M

Here is the rule working in the real world. Catton and Tondelstrand is an 18-person family law firm on the Sunshine Coast. When we started, we did not try to transform the whole firm. We looked for the one boring, repetitive task that ate the day.

It was email filing. Every incoming email had to be read, matched to the right matter, and filed in the right place. A 30 to 45 minute job, done across the team, every single day, that nobody wanted. A textbook first win.

So we pointed AI at that one task. It now reads every incoming email, works out which matter it belongs to, and files it automatically, around 4,000 emails a month with no one dragging and dropping. That first win earned the trust to do more. The firm is recovering $1.3M a year in productivity across the systems we have built since. It started with one task that took 30 to 45 minutes a day.

You do not need a big AI strategy to start. You need one boring task nobody wants to do, done well.

Why starting small matters

It is tempting to skip the small win and go straight for the ambitious project. That is the fastest way to lose your team and break a process. AI amplifies what you are already doing. Point it at a solid task and you amplify success. Point it at chaos and you just get faster chaos.

The small win does two things at once. It proves the return, so you know AI is worth the effort before you spend more on it. And it builds the team's trust, so the next change is met with curiosity instead of fear. Start small, build trust, then build bigger. The 30-45 Rule is the first step of that, not the whole journey.

AI is not a silver bullet, and most of the tools being sold right now are ChatGPT with lipstick. But pointed at the right repetitive task and built into how your business already runs, it changes what your day looks like. The first win is where you find that out.

Your next move

You do not need a tool or a strategy to start. You need the task. So watch your own week. The next time you catch yourself doing something repetitive that takes 30 to 45 minutes and that you would happily hand to someone else, write it down. That is your candidate.

From there it is a short step to working out what it costs you and whether AI should run it. If you want a hand finding yours, see how we help or start with where AI fits. When you are ready, book a 20-minute AI Profit Scan and we will help you find the first task worth handing over.

Frequently asked questions

What is the 30-45 Rule?
The 30-45 Rule is a simple way to find where AI should start in your business. Find a task that takes you 30 to 45 minutes every day, or most days, that you do not enjoy and that follows a pattern. That is your first AI win. It is frequent enough that automating it pays back quickly, small enough that you can trust a new system with it, and repetitive enough that AI can do it well.

Why is 30 to 45 minutes the sweet spot for a first AI task?
Because it balances payback against risk. A task that small feels harmless, so you are happy to hand it over, but it happens often enough that the saved time adds up fast across a week and a year. Anything much smaller is not worth automating, and anything much bigger usually carries too much judgement to trust to a new system on day one. The 30 to 45 minute daily task is the safest place to prove the return.

What kinds of tasks fit the 30-45 Rule?
Repetitive, pattern-based admin work. Common examples are email triage and filing, writing up file notes after meetings, generating quotes from a standard template, and drafting first-pass reports. The common thread is that the task follows the same steps every time and a person still checks the output before it goes out. That pattern is what AI handles well.

Why treat AI like a junior employee?
Because AI is like a junior employee you just hired. It is clever, tireless and eager to help, but you have to show it the ropes one job at a time and check its work at the start. You would not give a new hire the whole business on day one, so you give AI one clear, repetitive task, build trust as it proves itself, then expand. AI amplifies what you are already doing, so you point it at a solid task first.

Does the 30-45 Rule actually work?
Yes. Catton and Tondelstrand, an 18-person family law firm on the Sunshine Coast, started with one 30 to 45 minute daily task: filing emails into the right matter. That first win built the trust to expand into more systems, and the firm is now recovering $1.3M a year in productivity. It started with one boring task nobody wanted to do.

What is the first step to finding my AI win?
Watch your own week. The next time you catch yourself doing something repetitive that takes 30 to 45 minutes and that you do not enjoy, write it down. That is your candidate. You do not need a strategy or a tool first, just the one task. From there you can size what it costs you and decide whether AI should run it.