The Three-Bucket AI Framework
Three questions that stop a management team from getting lost in AI technology.
Most companies begin their AI journey by asking a single question: where can we become more efficient with AI? That question leads immediately to ChatGPT licences, Copilot rollouts, prompt training, HR automation, reporting and document creation. These are real opportunities. They are also one third of the picture.
A management team should ask three fundamentally different questions instead.
01 Product — AI in what we sell
How does AI change our product, our service or our business model? Can it improve what we already sell? Can new features emerge? Can we create new services from data we already hold? What happens if a competitor puts AI into their offering before we do?
This is the most strategic bucket, and the one most rarely examined. Its economic effect is growth and differentiation: revenue, customer value, business model.
02 Production — AI in how we create it
How does AI change the way we create our product or service? This is the actual value-creation process: software development, engineering, research, design, manufacturing, quality assurance, proposal creation, documentation.
The point management usually misses is that AI does not only automate administration. It can change the economics of producing the product itself. Its effect is speed, output, quality and cost of production.
03 Organisation — AI in how we run the company
How does AI change the way we run the organisation? Finance, HR, sales, marketing, legal, administration, management reporting. This is where most companies start, because it is the easiest place to start. Its effect is efficiency and overhead.
The second dimension: Replace, Augment, Create
Every opportunity in every bucket can be read in one of three modes. Replace: can AI take over the task entirely? Augment: can AI make a person substantially more capable? Create: can AI enable something that was not economically or technically sensible before?
Three buckets by three modes gives a nine-field canvas. That is the working tool. It fits on one page and it survives a management meeting.
The asymmetry
Organisation is often the easiest to implement. Production often carries the larger productivity impact. Product often carries the largest strategic impact. This is not a law, and every company is different. But the balance frequently shifts that way — which means most companies spend their attention in the least strategic place.
Why it matters
Most companies ask how AI makes their employees more productive. The more important question is how AI changes what their customers will pay for tomorrow.
This is not an academic theory. It is three simple questions I use in workshops to keep management teams out of the technology weeds — and it is where every AI Opportunity Scan I run begins.