Why “AI my business” is a red flag

Looking for support with Building a Business Operating System of your own – see my Business Operating System Page

Get the first chapter of my book: AUTOMATED – The Operating System for the One-Person Business

A business owner in the US once asked me to “AI my business”.

I turned the work down.

Not because I am against AI. I build AI and automation systems for businesses. I use AI in my own work, I help clients find practical automation opportunities, and I think the right AI setup can save a business a serious amount of time.

But “AI my business” is not a brief.

It is a red flag.

It usually means the person can feel the pressure, but has not yet diagnosed the problem. They know the business is too manual, too slow, too dependent on people remembering things, or too messy behind the scenes. They know something needs to change.

But they have jumped straight to the tool before understanding the workflow.

That is how AI projects go wrong.

AI can help a business. But it cannot tell you, on its own, which part of the business is worth fixing. It cannot decide whether your real problem is sales follow-up, onboarding, quoting, reporting, task management, delivery, documentation, customer support, or knowledge management.

And if the underlying process is unclear, AI usually makes the confusion faster rather than better.

That is why the first question should not be:

“How do I AI my business?”

The better question is:

“Where is my business leaking time, money, decisions or attention, and which of those problems is actually suitable for AI or automation?”

That is a much more useful place to start.

It is also the reason I created my Systems Diagnostic. Not as a way to sell AI for the sake of it, but as a way to review the business properly, identify the best opportunities, and build a practical roadmap before anyone wastes money on the wrong thing.

“AI my business” sounds modern, but it does not mean anything

I understand why people say it.

AI is everywhere. Every software platform has added an AI feature. Every consultant is talking about agents, automation, copilots, prompts and workflows. If you run a business and you are already stretched, it is natural to think:

“There must be a way to use AI to take some of this off my plate.”

That instinct is right.

The wording is the problem.

“AI my business” is too vague to act on. It does not tell me what hurts. It does not tell me where the time is going. It does not tell me what the business does repeatedly. It does not tell me where the bottlenecks are. It does not tell me what data exists, where it lives, or whether anyone trusts it.

It is the same as saying:

“Can you technology my business?”

Or:

“Can you systemise everything?”

It sounds like a request, but it is really a signal that we need to slow down and diagnose the business first.

Because AI is not a strategy. It is not a business model. It is not a process. It is not a substitute for knowing how the work actually happens.

AI is a tool.

And tools only work when they are pointed at the right problem.

What people really mean when they say “AI my business”

When someone asks me to “AI my business”, they usually do not literally mean they want AI everywhere.

What they really mean is one of these things:

“I am tired of being the person everything runs through.”
“I want the business to keep moving when I am not watching it.”
“I want fewer decisions landing on my desk.”
“I want my team to stop asking me the same questions.”
“I want sales follow-up to happen properly.”
“I want client onboarding to stop being so manual.”
“I want information to be easier to find.”
“I want fewer things slipping through the cracks.”
“I want to grow without adding more admin, more hours, or more people.”

Those are all valid problems.

But they are not all AI problems.

Some are workflow problems. Some are data problems. Some are ownership problems. Some are process problems. Some are communication problems. Some are “you do not have a system of record” problems.

AI might be part of the answer, but it might not be the first part.

Sometimes the better fix is a form, a checklist, a CRM stage, a task template, an automated email, a better handoff, a clearer owner, or a proper Business Operating System.

This is why I am cautious when someone starts with AI.

Not because AI is useless. It is not.

But because if you start with AI before you understand the business, you are very likely to build the wrong thing.

AI amplifies what is already there

This is the bit people often miss.

AI does not magically create a calm, organised business. It amplifies the business you already have.

If your workflows are clear, your data is consistent, your team knows where the truth lives, and your processes are reasonably standardised, AI can be extremely useful.

It can draft follow-up emails. It can summarise calls. It can route requests. It can create first drafts of proposals. It can classify support tickets. It can prepare reports. It can turn messy notes into structured actions.

But if your workflows are unclear, your data is scattered, and every task depends on someone remembering the “real” way things work, AI will amplify that too.

It will create plausible but unreliable output.

It will pull from the wrong source.

It will automate a process nobody has agreed on.

It will send work to the wrong place faster.

It will give you more moving parts to maintain.

That is not progress. That is faster confusion.

This is why I often say that AI should come after the business has been made legible. Before you automate a workflow, you need to understand it. Before you ask AI to make decisions, you need to know what good judgement looks like. Before you connect tools together, you need to know where the truth lives. Before you build an AI agent, you need to know what the agent is allowed to do, what it must not do, and where a human needs to approve the output.

That is the work most businesses want to skip. It is also the work that determines whether the AI project succeeds.

The wrong order: tool first, diagnosis later

The most common mistake is starting with the tool.

Someone sees a demo. They hear about ChatGPT, Claude, Make, Zapier, Notion AI, Pipedrive AI, Microsoft Copilot, or some new agent platform. Then they start looking for places to use it.

That is backwards.

The wrong order looks like this:

Pick an AI tool.
Look for something to use it on.
Build a workflow around the tool.
Discover the data is messy.
Realise the process is not clear.
Add manual workarounds.
Lose trust in the system.
Quietly stop using it.

That is how businesses end up with half-finished automations, duplicated records, broken handoffs and a team that does not trust the new system.

The better order is:

Diagnose the business.
Find the recurring workflows.
Identify the bottlenecks.
Decide what outcome matters.
Clean up the process.
Choose the right tool.
Add AI only where it actually helps.
Measure whether it worked.

That is a very different conversation.

And it starts with a better question.

Not “How do I AI my business?”

But:

“Where is the business stuck, and what is the simplest reliable way to fix that?”

The real problem is often dependency on one person

In many small businesses, the biggest issue is not the lack of AI.

It is that the business depends too heavily on one person.

Sometimes that person is the owner. Sometimes it is the operations manager. Sometimes it is the best salesperson, the project lead, the senior administrator, or the person who has been there long enough to know how everything really works.

They are the person everyone asks.

They know where the files are.

They know which client needs special handling.

They remember what was agreed on the call.

They know how the quote should be prepared.

They know which supplier to chase.

They know which spreadsheet is current.

They know the workaround nobody documented.

That person becomes the operating system.

And that is dangerous.

Because when the business depends on one person’s memory, attention and availability, it cannot scale properly. It can only move as fast as that person can think, answer, chase and decide. This is the deeper problem behind a lot of “AI my business” requests. The owner does not really want AI. They want the business to stop depending on them for everything. Or they want their key person to stop being the bottleneck. That is the transformation I care about.

I turn businesses that depend on one key person into businesses that run on systems.

AI can be part of that. Automation can be part of that. Tools like Notion, Pipedrive, Asana, Airtable and Make can be part of that, but the tool is not the transformation. The transformation is moving the knowledge, decisions and repeated work out of someone’s head and into a system the business can actually run on.

I learned this in my own business too

I have been working with businesses for more than 25 years, across more than 400 clients. For a long time, I built my own consultancy the way many people build theirs. More clients, more calls, more projects, more responsibility, more things depending on me. Eventually, I had to admit that the issue was not just workload. The issue was dependency.

Too much of the business still depended on me knowing what to do next, remembering the right detail, following up at the right time, and keeping the whole thing moving manually. So I did to my own business what I now do for clients.

I made the work clearer. I built systems around it. I created a Business Operating System. I moved recurring decisions, tasks and follow-ups into structured workflows. I used automation where it made sense. And only then did AI become genuinely useful.

That is why I am sceptical of vague AI projects.

Not because I am anti-AI.

Because I know what happens when you try to automate a business that has not been made clear yet.

You do not get freedom.

You get a faster version of the same mess.

When AI is useful

AI is useful when the job is clear.

For example, AI can help with:

Drafting sales follow-up emails from structured call notes
Turning discovery notes into a proposal outline
Summarising meeting transcripts into actions
Classifying inbound requests
Creating first drafts of standard documents
Extracting key details from forms or emails
Preparing weekly reports from project updates
Suggesting responses to common support questions
Repurposing content into articles, newsletters or social posts

Those are useful because the input, output and purpose can be defined.

The AI is not being asked to “fix the business”.

It is being asked to perform a specific role inside a clear workflow.

That distinction matters.

“Use AI to summarise every sales call and create the next follow-up task in Pipedrive” is a brief.

“Use AI to review support requests, classify them by urgency, and draft a suggested reply for human approval” is a brief.

“Use AI to turn a completed project form into a client onboarding checklist” is a brief.

“AI my business” is not.

The more specific the workflow, the more useful AI becomes.

When AI is not the right first step

Sometimes the answer is not AI.

If your CRM stages are inconsistent, fix that first.

If your team does not use the same task system, fix that first.

If your onboarding process changes every time, document the baseline first.

If project information is scattered across email, WhatsApp, Google Drive and someone’s memory, create a system of record first.

If nobody agrees what “done” means, define that first.

If every client is treated as a special case, standardise the core process first.

If you cannot explain the workflow in plain English, you are not ready to automate it.

This is not glamorous work, but it is the work that makes AI useful later.

In many cases, the highest-value fix is not an AI agent. It is a clean workflow, a proper intake form, a better CRM setup, a project template, a handoff checklist, or a single place where the truth lives.

That may sound less exciting than AI.

But it is usually what saves the time.

The better question to ask

Instead of asking:

“How do I AI my business?”

Ask:

“Where is the business leaking time, money or attention?”

Then get specific.

Where are people doing the same manual task every week?

Where are customers waiting too long?

Where does work get stuck?

Where do people ask the same questions repeatedly?

Where does information go missing?

Where are errors happening?

Where are decisions being delayed?

Where are you personally still in the loop when you should not need to be?

Where is a key person carrying knowledge that should live in the system?

Those questions lead somewhere useful.

They show you whether the answer is AI, automation, process design, documentation, better task management, better CRM setup, or a proper Business Operating System.

They also help you avoid building something impressive that does not solve the real problem.

This is why I created The Systems Diagnostic

The reason I created The Systems Diagnostic is simple.

I do not want to sell someone an AI build until we know what problem we are solving.

The Systems Diagnostic is a review of your business workflows, systems and automation opportunities. It is designed to answer the question people should be asking before they ask for AI:

“What is actually worth fixing, and what is the right way to fix it?”

In the diagnostic, I look at things like:

Which workflows repeat every week
Where time is being lost
Where handoffs are breaking
Where information lives
Which systems are trusted
Which decisions need human judgement
Which tasks are safe to automate
Where AI could genuinely reduce work
Where simple automation would be better than AI
What should not be automated at all
What needs to be standardised before anything is built

The output is not a vague AI strategy.

It is a practical roadmap.

What to fix. In what order. What to automate. What to leave alone. Where AI belongs. Where it does not. And what the business needs before a build makes sense.

That is much more useful than rushing into an AI project because AI happens to be fashionable.

The right order: diagnose, systemise, then automate

The order matters.

1. Diagnose the business

Start by understanding how the work actually happens.

Not how the process is supposed to work. How it really works.

Where does work enter? Who touches it? Where does it get stuck? What information is needed? Who makes the decision? What happens when that person is away? What is repeated? What is bespoke? What is valuable? What is waste?

This gives you the map.

Without the map, you are guessing.

2. Build the operating system

Once you understand the business, you can create the system.

That might include a CRM, project management setup, Notion workspace, Airtable base, Asana structure, SOP library, intake process, reporting dashboard, or a combination of tools.

The point is not the software.

The point is that the business has a clear way to capture work, triage it, move it through delivery, track responsibility, store knowledge, and review progress.

This is where the business becomes legible.

3. Add automation and AI

Only then do you add automation and AI.

At this point, you know what the workflow is. You know where the data lives. You know what the output should look like. You know who approves what. You know what success means.

That is when AI can actually help.

Not as a magic layer sprinkled over the business, but as a specific tool inside a system that already makes sense.

A simple readiness checklist

If you are wondering whether your business is ready for AI, ask yourself these questions:

Can we describe the workflow clearly?
Do we know who owns it?
Do we know what good output looks like?
Do we know where the required information lives?
Is that information reliable?
Do we have consistent statuses, stages or labels?
Do we know which decisions require human judgement?
Do we know what should never be automated?
Can we measure whether the change worked?
Would this still work if the key person was away for a week?

If you cannot answer those questions, you probably do not need an AI build yet.

You need a diagnosis.

That is not a delay. It is how you avoid wasting money.

The real promise of AI and automation

The promise of AI is not that your business becomes futuristic. The promise is that the right work happens with less chasing, less remembering, less waiting and less manual effort.

A lead gets followed up.

A client gets onboarded.

A task gets created.

A proposal gets drafted.

A report gets prepared.

A handoff happens cleanly. A question gets answered from the right source. A key person is no longer the only one who knows how something works. That is what matters.

Not whether the solution sounds impressive.

Not whether you can say you have an AI agent.

Not whether a tool demo looked clever.

The question is whether the business runs better afterwards.

Do not AI your business. Diagnose it first.

If you are thinking, “I need to AI my business”, pause for a moment.

You might be right that AI can help but the first step is not picking a tool, hiring someone to build an agent, or asking ChatGPT to write a process for you. The first step is working out where the business is actually stuck.

Where is time leaking?

Where are people waiting?

Where are decisions delayed?

Where is the business too dependent on one person?

Where would automation create real leverage?

Where would AI create risk?

Where would a simple system fix the problem faster?

That is the work that needs to happen first.

If you want a quick starting point, use my Free Business Automation Audit. It will help you identify where your business is losing time and where automation might help. If you want a deeper review, that is what The Systems Diagnostic is for. It is a review of your workflows, systems and automation opportunities, designed to give you a practical roadmap before anything gets built.

Because “AI my business” is the wrong question.

The better question is:

“What needs to change so this business runs better, with less dependence on one person?”

Answer that first. Then AI has a job worth doing.

Prefer to talk it through ?

Book a Free Scoping Call or take a look at the main Business Operating System Consultant page.

Follow Me

Why “AI my business” is a red flag