3 signs your Brisbane business is ready for AI (and 2 that say wait)

I get asked constantly whether a business is 'ready for AI' and the honest answer is that the size doesn't matter, the industry doesn't matter, but three specific things really do. Miss any of them and the AI investment burns money for six months before you cancel it.

01The three green lights

1 · You've written down at least one process

It doesn't have to be a manual. It can be a page of dot-points. But somewhere in your business, there's a documented version of how at least one thing gets done — how a lead becomes a quote, or how a job becomes an invoice, or how a client complaint gets handled.

This matters because AI is a process amplifier. If the process only lives in the owner's head, the AI can't be trained to do it consistently. If there's a written version — even a rough one — the AI has something to learn from.

The businesses we've seen struggle most are the ones where every job is bespoke and nothing is written down. AI can't help there until you write down at least the 20% that repeats.

2 · Your data lives in tools, not paper

Xero for money. Some kind of CRM (HubSpot, HighLevel, ServiceM8, Tradify). A calendar. If those three tools cover 80% of your operational data, you're ready. If half your business runs off a physical diary or a whiteboard in the office, AI has nothing to read.

You don't need everything digital — one of my best clients still runs their job-sheets in a paper diary. But the money, the leads, and the schedule need to be in tools with an API.

3 · You have >2 hours/week on something repetitive

AI investment pays back where the repetitive task is expensive enough to matter. Rough rule: if you can point at one job (invoice-chasing, quote follow-up, job-brief formatting) that costs you >2 hours a week manually, the AI investment usually pays back inside 12 months.

Below 2 hours it's often not worth the setup cost. Above 5 hours it's almost always worth it.

02The two red lights

1 · Cash flow is stressed

AI investment during a cash squeeze is a bad idea. The build cycle is 4–8 weeks and the value comes 2–3 months after that. If you need money back in the door this month, spend the AI-project money on marketing, not automation.

This is the most-common mistake we talk owners out of. AI is a good investment in a healthy business, not a rescue for a sick one.

2 · You're the only one who knows how anything works

If the business runs entirely out of your head — nobody else knows the client list, the pricing rules, the process — then AI can't be your first project. Your first project is documenting the business so a bus doesn't take the whole company with you.

This is the boring answer nobody wants. Six months of writing things down, hiring one person to run the operational bits, and then commissioning AI to make that written-down business run more smoothly. Do the boring work first — the AI is easy afterwards.

03What 'ready' actually looks like

The Brisbane SMBs where AI has genuinely worked shared a pattern:

• 3–20 staff.
• Xero for money, some kind of CRM/job-management platform, and one shared inbox.
• At least one owner-documented process (even a rough one).
• Consistent revenue for the last 12 months (not necessarily growing, but not tanking).
• A specific pain — 'I spend Tuesday nights on invoice chases' — not a vague ambition to 'be more AI'.

If that describes you: get a Leak Check. If it doesn't: fix the gap first, then come back.