A client came to us ready to buy. They had the budget, they had decided, and they wanted an AI system to answer customer enquiries and track their orders. It would have been an easy sale.
We told them not to buy it. Here is why.
What we found#
When we looked at how the business actually ran, the picture was clear. Their orders lived in a paper book and a WhatsApp chat. Their stock was counted by eye. Nothing the AI would need to answer from was written down anywhere a computer could read.
They were at stage 0 of the ladder - and the system they wanted to buy was a stage 3 system. There were three missing rungs between where they stood and what they were about to pay for.
Selling them that system would have been selling them a failure. It would have been built on a foundation that was not there.
What we told them instead#
We told them to keep the AI budget in their pocket and do one cheaper thing first: get their orders and stock into a simple digital system. That is it. A fraction of the cost, valuable on its own - and the thing that would make an AI project actually work later.
They were surprised. They had come expecting to be sold to. Instead they left with a smaller bill and a step that made sense.
Why this is the whole point#
An AI system built on a business that is not written down does not half-work. It fails, publicly, in front of customers - and then the owner concludes that "AI does not work for a business like mine." The truth is that the AI was fine. The foundation was missing.
A partner who sells you the top of the ladder while you are standing on the bottom is not doing you a favour, whatever the demo looked like. The most useful thing we can sometimes say is not yet - and here is exactly what to do first.
That client will buy the AI system eventually - and when they do, it will work, because the rungs beneath it will be there. For the full map of the climb, see the five stages of AI adoption.