Most marketers work with Claude in a chat window, one conversation at a time, and leave the rest sitting there. Twenty questions about how you actually work today, in plain English. At the end: where your setup stands, the biggest thing you are missing, what it is costing you, and exactly what to do today, this week and this month.
Ten questions about what actually happens: what Claude knows about your business, whether good work repeats, whether anything carries forward, how it fits your week, what it is allowed to do, and how the work gets checked. This is what produces your result.
Ten things Claude can do, each described by what it would do for you rather than by its name. This never adds to your score. Knowing every feature is not the same as running a good setup, and a good setup does not need every feature. It is used for one thing: naming what is worth your attention next.
Every question works the same way. Five options, arranged from earliest to furthest along. Pick the one that describes how you work today, not the one you are aiming for. The honest answer produces the useful result, and nobody is watching.
Not what you plan to set up. Not what you saw someone describe on LinkedIn. Ten questions across six things that decide whether a setup holds together. The ones that matter most get asked twice, from two angles.
Ten things Claude can do, named by what they would do for you. The product names are there too, in small type, so you can go and find them. There is no prize for the far end of the scale. Several of these are worth adopting late, or not at all.
Each one is a complete step, not a stage of a bigger project. Do the first one and stop, and you are still better off than you were this morning.