Seven questions plus one compliance gate. You get a maturity band, the two gaps to close first, and how to run the tests you are missing.
0% 7 to go
Does the designed journey reach its outcome?
Does every designed exit fire, and carry context?
Does it surface the information it should, and nothing it should not?
Can the customer step off the path and be put back on it?
Does it hold for the customer who does not cooperate?
Does the same input produce the same outcome every time?
Does it invent, drift, break persona or expose data?
Does your chatbot or voicebot serve customers in the EU?
This is a gate, not a scored question. Disclosure is required by law in the regions you deploy into, so a failure here stops a deployment rather than lowering your band. It is not weighted and does not affect your result.
