The preparation of the student is extremely important. I tutored students in R until last spring -- the students were in a PhD program that had to learn how to use data, but none had a programming class beforehand. Most were ok, but the last two had huge hurdles (one lost a job and was in the middle of a divorce). All were in a public policy program, so not having a precursor for learning data models hindered, though not impossible. I think the rise of AI as a programming buddy will alter how students seek help and can make tutoring mute in many instances.