Questions that produce something usable
Do not ask what should be in the course.
Asking an employer what should be taught invites them to speculate about curriculum design, which is not their expertise, and produces either a list of software names or a general statement about attitude and communication. Both are difficult to act on.
The productive questions are about their experience. Describe the last graduate you took on. What could they do on day one. What took longest to bring them up to speed. What did they arrive believing about the job that turned out to be wrong. Each of these produces specific, actionable material.
The question about mistaken expectations is particularly valuable and rarely asked. It surfaces where the provider's description of the industry has drifted from reality, which affects recruitment and student satisfaction as much as curriculum.
It is also worth asking what they would keep. Consultation focused entirely on deficiency produces a distorted picture and can lead providers to abandon things that were working.
There is one question worth adding to any employer conversation, which is what they would want a graduate to be able to do that nobody currently teaches. It surfaces emerging requirements before they appear in any framework, and employers frequently have a clear answer that they have never been asked for.