There is more noise about AI than at any point in the last decade. Every newsletter, every conference, every LinkedIn feed is telling small business owners that they need to act now or be left behind. Some of that pressure is legitimate. A lot of it is engineered to sell tools and consultancy.
What is also true: businesses that build genuine AI capability into their workflows now — methodically, with clear thinking about what they need and why — are accumulating an advantage that compounds. Their people are learning, their processes are improving, their costs are falling relative to competitors. The businesses that delay are not standing still. They are losing ground against a moving target.
The problem is that “methodically” is hard when you don’t know which tools are worth using, how to evaluate whether AI will actually help your specific operation, or how to justify the cost. That is exactly what I can help with.
What I actually am
I am not an AI expert. I have no background in machine learning or data science, and I will not pretend otherwise. What I have is daily, practical experience using AI tools across a wide range of real business tasks — writing, analysis, research, automation, client work — and the methodological background to apply that experience usefully. My job is not to sell you on AI. It is to find out whether AI will help your business, by how much, and what to do first.
How it works
Each engagement begins with a conversation and ends with a written set of prioritised, costed recommendations. In between:
- Discovery conversation. Before anything else, I need to understand your business — what it does, how it actually runs, where time goes, what frustrates you, what you want to grow or protect. This is not a sales call.
- Workflow and process audit. Using established process mapping techniques, I document your current workflows and identify tasks that are repetitive, rule-based, time-consuming, or produce variable results — the characteristics that make a process a candidate for AI support.
- AI-applicability assessment. Not every task that could be assisted by AI should be. I evaluate each candidate against three criteria: the realistic capability of current tools to handle it, the effort required to implement and maintain a solution, and the risk if the AI gets it wrong. This produces a prioritised list, not a wishlist.
- Tool research and matching. I research and, where appropriate, test specific tools against the tasks identified. I have no commercial relationship with any AI platform or vendor.
- ROI modelling. For each prioritised recommendation, I build a straightforward cost-benefit model: the cost of the tool or implementation, the realistic time saving, the value of that time, and the payback period. The model is yours to keep and update.
- Pilot design. The highest-value recommendation becomes a structured pilot — a low-risk, time-bounded test with clear success criteria. I help design it and, if needed, support the implementation.
- Governance and data risk review. Before any AI tool touches real business data, I check what that tool does with it, what the GDPR and confidentiality implications are, and whether there are risks you should know about. This is not optional.
- Follow-up assessment. After the pilot, we measure actual results against the projected ROI. If the projection was wrong, we understand why. That learning is part of the service.
What I won’t do
I won’t tell you AI will transform your business when the honest answer depends entirely on what your business actually does. I won’t recommend tools I earn commission from — there are none. I won’t design a programme intended to eliminate your staff. I won’t stay in an engagement longer than it needs to be. And I won’t give you a report that could apply to any business: the value of this process is that it is specific to yours.
The window is real
The businesses building genuine AI capability right now are not the ones buying the most subscriptions. They are the ones doing it with clear thinking about where it actually helps, what it actually costs, and how to make it actually work. That institutional knowledge — what their specific workflows respond to, what their team can adopt, what produces real return — is not something you can buy later. It compounds from the moment you start building it.
The right response to the pressure you are feeling is not to act fast. It is to act with enough structure to make sure what you do works. That is what this service is for.