Ask any teacher and they'll tell you the same thing: most students look at the grade, glance at the comments for a second, and move on. That's not a student problem — it's usually a feedback problem, and it's fixable.
Lead with the grade's opposite
If the grade comes first, it becomes the whole story and the comments become an afterthought. Where you can, put the grade at the end of your comments instead of the top, or at minimum don't lead with it. Students read differently when they don't already know the verdict.
Point at one thing, not everything
A page covered in corrections feels like a punishment, not a lesson, and students respond to it by skimming past all of it. Pick the one or two things that matter most for this piece of work and go deep on those, even if you noticed six other issues. A student who fixes one real problem next time learned more than one buried in twelve minor notes.
Make the next action obvious
"Weak transitions" describes a problem. "Try starting paragraph 3 with a sentence that connects back to your thesis" describes an action. The second version is the one a student can actually do something with tomorrow, not just feel bad about today.
Separate what's fixable now from what's a longer-term goal
Some feedback applies to the next revision. Some feedback applies to the next essay, or the next unit. Being explicit about the difference — "revise this before Friday" versus "keep working on this across the semester" — stops students from either ignoring long-term feedback or panicking about short-term feedback they can't act on immediately.
Ask for a reply, at least sometimes
A simple "reply with one thing you're going to try next time" turns feedback from something that happens to a student into something they have to process. It doesn't need to happen on every assignment, but doing it occasionally noticeably increases how seriously comments get read the rest of the time.
Where AI actually helps here
The bottleneck for detailed feedback usually isn't knowing what to say — it's having time to say it to 30 or 150 students individually. An AI grading tool built around your rubric can draft specific, text-referencing feedback for every submission, which turns your job into editing and personalizing rather than writing from a blank page each time. That's the gap ClassGrade Pilot is built to close: you keep control of what actually gets sent, but you're not typing the same core comment two dozen times.
Feedback that gets read is usually shorter, more specific, and more actionable than feedback that gets ignored — not more thorough.