Article · Getting started with AI

    Should I fix my processes before I start using AI in my business?

    By Nina Huchthausen, Strategic AI Partner & Trainer · Published

    "Do I need to sort out my processes before I use AI?"

    I get asked some version of this almost every week. Sometimes it comes in a discovery call, sometimes it comes at the end of a workshop when everyone is packing up their laptops, and sometimes it comes as a slightly guilty confession: "Nina, I think our processes are a bit of a mess, so maybe we're not ready for AI yet."

    If that is you, you are in very good company. And the short answer is yes, but probably not in the way you think. You do not need perfect processes. You need to understand how work actually gets done in your business, because AI will copy whatever you give it, and that includes the mess.

    Why does AI make a messy process worse, not better?

    AI speeds up whatever process you give it, including the broken parts. If work in your business relies on people remembering things, chasing each other and copying information between five tools, AI will not fix that. It will do the same chaos faster. Untangle the process first, then train AI on the version that works.

    I have been working with processes and systems for over 16 years, and the pattern I see right now is a business owner who feels behind on AI, signs up for a shiny new subscription, points it at the busiest part of the business and hopes for the best. And what they end up with is a high-speed hot mess.

    Because if your business is running on bottlenecked people and systems held together by spreadsheets, sticky notes and the contents of your brain, AI does not know which of those steps matter and which ones are just habit. It does exactly what you tell it to, including the steps that should never have been there in the first place.

    The research backs up how much of the working day this mess is eating. Asana's global study of more than 9,600 knowledge workers across six countries, including Australia, found that 58% of the working day goes on "work about work", which is the time spent coordinating work rather than doing the skilled, strategic jobs people actually want to do.[1]

    That is more than half the day spent on coordination. Now imagine handing that coordination to AI without first asking why it exists. You would be paying for a very clever tool to keep doing the work that should have been designed out.

    So the question is not "how can we use AI in our business?" The better question is the one I ask every client in our first session: why the heck are we still doing it this way?

    How do I know if my processes are the real problem?

    Run a simple five-point reality check on one process, from start to finish. Follow the work, find where it waits on a person, spot the jobs your team repeats every week, question your tools and ask what would stop if you disappeared. If those answers make you wince, your processes need attention before AI does.

    This is the check I walk business owners through before we go anywhere near building an AI assistant. Pick one process. Just one. Quoting, onboarding a new client, month-end, whatever is keeping you up at night.

    1. Follow the work. Walk that process from start to finish. How many steps does it actually take? Who touches it? Where does it get stuck?
    2. Find the human bottlenecks. What cannot move forward without someone answering a question, approving something or finding information? Is that person you?
    3. Spot the repeat offenders. What is your team copying, pasting, rewriting, chasing or manually updating every single week?
    4. Question the system. Are you using five different tools to do something that could happen in one? Are your people working around your software instead of with it?
    5. Ask the holiday question. If you disappeared for two weeks tomorrow, what would grind to a halt?

    That last one is the one that tends to land. One owner I worked with this year had built a business over many years and still could not take three weeks off. The business could not run without them.

    When owners do this exercise, they nearly always find the same thing. The process itself is not complicated. It has just grown sideways over the years, one workaround at a time, until nobody remembers why there are seven steps when two would do.

    What do I sort out first: people, processes or systems?

    Start with processes, then people, then systems, then bring in AI. Get clear on how work really happens, make sure your team knows who owns what, and give information one home. AI becomes the fourth resource on top of those three, not a replacement for any of them.

    Every business runs on three things: people, processes and systems. When all three are carried by your people, including the manual, repetitive, backstage work that keeps the machine turning, that is when everything ends up on your desk.

    Here is the order I work through with clients, and what each one looks like when it is sorted.

    1. Order 1ProcessesYou can describe how work really gets done, not how you think it gets done.
    2. Order 2PeopleYour team knows who is responsible for what and feels confident making decisions without running to you for permission.
    3. Order 3SystemsInformation and decisions live in one place, not in 47 spreadsheets, six WhatsApp groups and someone's head.
    4. Order 4AIYou train AI on how your business works, so it can carry the backstage work with a human checking the result.

    Notice that AI comes last, and that is on purpose. When we train AI on your people, processes and systems, AI becomes the fourth resource. It takes on the Groundhog Day work, and the work that makes your business great stays with humans.

    This does not have to be a twelve-month project before you are allowed to touch AI. In the businesses I work with, we start training AI from day one. We pick one small process, map how it really works, tidy it up and train AI on it, all in the same day. Keep it small, get it working, then move on to the next one. Bite by bite, not big bang.

    When is my business ready for AI to take work off my team's plate?

    You are ready when you can describe one process clearly enough that a new staff member could follow it. That same clarity is what AI needs. You do not need every process perfect, just one well understood process, a person who owns it and a human who checks what AI produces.

    I always say that AI starts like an intern. Incredibly smart, studied a million subjects, but it has never been inside your business before. You would never hire a new staff member, give them no training and no direction, and then wonder why they are doing a sh.t job. AI is exactly the same.

    So the readiness test is simple. Could you sit a new team member down tomorrow and walk them through this process, step by step, without saying "oh, it depends" every two minutes? If yes, you are ready to train AI on it.

    Here is what it looks like when it works. One business I work with had a reconciliation that took about an hour every time it ran. Once we had cleaned up the process and trained AI on it, the same job took around 15 minutes, and the person responsible checked every line before signing it off. AI did the first pass. The human stayed on top.

    That is also why most small businesses do not need autonomous AI agents yet.

    That is the part I care about most. Training AI is not about handing your business over to a robot and hoping for the best. It is about giving your people their best hours back, so they can spend them on the work that grows the business and still make it home for dinner.

    Asana's research found that knowledge workers believe better processes alone could save them 4.9 hours every week.[1] Train AI on those better processes, and you start to see where the real capacity comes from.

    Where do I start fixing my processes?

    Not with another subscription. Start with one process, one honest look at how it really works, and one question: what should be simplified, what should be handed to AI, what absolutely needs a human, and what should not be happening at all?

    Because the goal is not to build the most technologically impressive business in town. It is to build a business that works brilliantly, grows sustainably and does not need you to be the human glue holding the whole thing together.

    If you want to go one step further back, my article on where to start with AI and how to claim the work you love helps you decide what AI should and should not go near in the first place.

    Fix the process. Then train the AI. That is where the real magic lives.

    "But I can't even explain one of my processes." That is okay.

    You do not need to work that out on your own. That is where I come in. We never look at your whole business at once. We solve one bottleneck at a time, and it always runs the same way.

    1. We sit down together, talk it through, and I ask you all the right questions.
    2. I map the process quickly, and you approve it.
    3. Together we decide what goes to AI and what stays with you.
    4. We build it together, I teach you how it works, and I stay alongside you until the new way of working feels easy.

    Then we move on to the next one. Small and steady, one process at a time.

    Get in touch and tell me which process is driving you mad.

    Not sure which task to hand over first?

    My free AI Opportunity Map works a little like an AI interviewer. It walks through your business with you and shows where AI could safely make the biggest difference for you, and how much time you could get back.

    Get it here 👉 Your AI Opportunity Map

    Would you rather talk it through? Let's have a chat. No prep. Just an honest conversation.

    Book a free 30-minute call

    Keep reading

    Nina Huchthausen, Strategic AI Partner & Trainer at Set to Growth

    Nina Huchthausen

    Strategic AI Partner & Trainer, Set to Growth

    Nina is a process engineer by trade. For 16 years she has helped businesses streamline their processes and systems, and now she trains AI to take the manual, repetitive work off their team's plate. Her clients feel the draining work lift, and get either the room to grow and scale, or the same revenue in a 7 hour day instead of 12.