Some points:

• This would work if both the parents and the students wanted advising for the students.
• Teenagers and young adults are often rebellious and don't want to do what their parents tell them to.
• If the students don't want advising, they won't benefit much from it.
• A large fraction of parents are willing to pay for SAT prep and such, but that's very different from what we're offering.
• If the children are sufficiently young, there's less of an issue of teenage rebelliousness, and parents are more directly involved in their children's education.
• The only parents who have made contact with us are parents of young children. (Edit: I misremembered – there were a few parents of older children who contacted us, but their children weren't interested.)
• The parents of young children who contacted us expressed willingness to pay much more often than the students who contacted us.
• Our expertise is much more relevant to high school and college students than to elementary and middle school students, except for the ones who are precocious to the point of being cognitively similar to high performing high school and college students. We haven't worked with other elementary and middle school students and have no background in early childhood development.
• As in my post, there's an issue of needing a large flow of clients (because the number of hours per client is small).

It could be that we haven't pursued this option in sufficient depth. We could to focus on elementary and middle school students, or see whether the parent-child correlation of interest in advising for older students is sufficiently high. We would guess that the correlation isn't high enough, though we only have a few relevant data points (examples or parents being interested when their children aren't - there are many examples in the other direction) and further testing could reveal otherwise.

Any thoughts?