The context ladder is a model I use in training: it shows how the quality of AI's answers rises step by step the more relevant context you provide – from the loose question, via role, purpose and target audience, to examples, constraints and the organisation's own material.
Most people start on the bottom rung of the ladder: a question without any context – and then judge the technology on the generic answer that comes back. But the model does not know who you are, who you are writing to, or what good looks like in your world, until you tell it.
The ladder makes prompting concrete without making it technical: each rung is a piece of context anyone can supply – and the top rungs, where the organisation's own knowledge and examples come into play, are also where the competitive advantage lives.
In practice
Try the difference for yourself: first ask a question on its own. Then give it a role, a purpose, a target audience and a good example from your own work – and compare the answers. That is the ladder, felt first-hand.
How to explain it to management
»AI is not a mind reader. The more of our reality we give it, the more of our reality it can help with.«
The context ladder is closely linked to the FOCUS model – and it is practised in all my courses and workshops.