Every week, I hear from businesses excited to implement AI, automation, or some other cutting-edge technology, and the first question I always ask is:
Have you taken a good, honest look at your processes first? More often than not, the answer is… not really.
Here’s the thing: AI doesn’t magically fix broken processes — it just amplifies what’s already there. If things are messy behind the scenes, AI is going to accelerate the chaos, not clean it up. If your operations aren’t well understood, layering AI on top will only create more confusion.
That’s why understanding your internal processes isn’t just a “nice to have”; it’s the foundation. Before you jump into AI, you need to know where you stand on the readiness scale.
Understanding the AI readiness scale
Across industries like manufacturing and mid-market enterprises, businesses are moving through different readiness levels:
- Basic awareness – recognising AI’s potential but having no structured processes.
- Operational readiness – eliminating inefficiencies, defining clear workflows, and integrating core systems (ERP, CRM, supply chain).
- Data maturity – ensuring high-quality, accessible, and structured data; AI is only as good as the data it learns from.
- Strategic AI planning – defining specific, measurable AI use cases rather than broad “we need AI” goals.
- AI-driven transformation – AI actively driving business value, improving decision-making, and optimising efficiency
Most businesses that make enquiries to us about AI are at level one or two. However, AI only starts to deliver real value at level four and five. If you’re not moving through these stages, AI will only create complexity, not efficiency. Ultimately, this disconnect is the underlying reason why statistics show that a frighteningly large number of AI projects fail to deliver. It’s not that the technology is bad — it’s just that the technical maturity isn’t there for it yet.
Five steps to take before considering AI
So, if you’re thinking about bringing AI into your business, here are five things to check off before we talk about it:
- Map out your business processes
You’d be surprised how many businesses don’t have a clear picture of how their operations run. They work in silos, with different teams doing things their own way. Before you layer in AI, take the time to map out key workflows — supply chain, customer service, finance, whatever’s core to your business. Where are the inefficiencies? Where does work slow down? If you can’t spot inefficiencies, AI is only going to make them faster and harder to unwind.
- Talk to your customers and end users
It’s easy to think you know what your customers want, but have you asked them lately? What frustrates them? What makes them come back? What would make their experience better? AI can help improve customer experience, but it needs to be pointed at the right problems. If you’re guessing, then you’re automating based on assumptions and fast-tracking to bad decisions at scale.
- Run workshops with your team
AI and automation don’t replace people; they enhance what your team is already doing. But if your internal processes are inconsistent or poorly understood, adding AI will just create more confusion. Bring your teams together. Sit down with your teams. Get their input on what’s working and what’s not. Identify pain points. Making sure everyone is aligned is critical before introducing automation.
- Get your data house in order
One of the biggest misconceptions that companies have is assuming that AI can “clean up” their data, but it doesn’t work that way. AI learns from your data, so if your data is full of gaps, errors, or inconsistencies, your AI models will be, too. Ensure your ERP, CRM, and supply chain systems are properly integrated and that your data is clean, structured, and accessible. Otherwise, AI will simply scale your existing data problems.
- Be clear on what you want AI to do
Many businesses approach AI with a vague goal: “We just want to be more efficient,” but that’s not a strategy. AI works best when applied to specific, measurable problems, such as predicting supply chain delays, reducing material waste, improving demand forecasting, or automating repetitive tasks. If you can clearly define the problem you’re solving, then you’re on the right path. If not, it might be worth spending a bit more time clarifying your goals first.
AI should never be the first step. It should be the final piece of a well-thought-out strategy. The companies that truly succeed with AI are the ones that deeply understand their own operations before implementing AI. They’re the ones who know their business inside out and use tech to build on what’s already working.
So, before you start the AI conversation, ask yourself: Have we done the hard work of understanding and optimising our business first?
If the answer is no, then AI is not your next step. I strongly recommend fixing your processes first, and AI will make them even better.
