Most feedback explains the grade. Good feedback changes the next piece of work.

By the time you reach the comment box, you have already done the work.

You designed an assessment that measures something real. You worked out the criteria, and what separates strong work from weak. You read sixty submissions and held the same standard on the last one as on the first. That is a serious amount of effort, and almost all of it is invisible to everyone but you.

Then all of it reaches the student as a number and one sentence: "Good work overall, but the analysis could be more critical." A sentence that would fit almost any paper, from almost any student, in almost any course.

The work was real. It just stops one step short of the person it was for.

The comment is the only part they receive

Everything before it happens on your side of the desk. The design, the criteria, the weighing of this paragraph against that one — the student sees none of it. They see a grade and a comment.

So the comment is where all that effort either lands or evaporates. Give it three vague lines and the student takes away a number, which is the one part of your work that tells them the least.

That's worth finishing properly, and finishing properly means something specific.

The comment that only defends the number

Look honestly at a set of comments you wrote last term and most of them will be doing one job: explaining the grade. The structure was unclear. Sources were thin. The argument didn't fully develop. All true. All pointing backwards, at work that has already been handed in and can't be changed.

The student reads it once, checks the number, and closes the file. Nothing about their next assignment is different because of it.

That's the honest test of feedback, and it has nothing to do with how much you wrote: could the student do something differently tomorrow because of this sentence? If not, the comment is defending a grade your criteria already justified — and everything you learned reading their work stays with you.

Feedback points forwards

The useful kind does the opposite. It still names what's in front of it — students need to know what you saw — but it spends its last sentence on the next piece of work.

Two comments on the same essay:

Your argument isn't fully developed and the sources are thin.

You make the case that the policy failed, but you use two sources and both agree with you. Next time, find one that disagrees and answer it — that's what turns a claim into an argument.

Same judgement. Same grade. The second one gives the student something to do on Monday.

It's also, and this matters at eleven at night, barely longer than the first.

Three things that make a comment usable

• Name what the work actually does. Not "the analysis is weak" — what is weak, in this

paragraph, on this question. A comment that would fit any submission tells the student nothing about theirs.

• Tie it to a criterion. If you graded "use of evidence", the comment belongs to "use of

evidence". When your comments and your rubric say the same thing in the same words, students stop asking where the grade came from.

• End with the next step. One concrete action, not a wish. "Be more critical" is a wish. "Find a

source that disagrees and answer it" is a step.

Three rules, and every one of them is easy at nine in the evening and hard at eleven.

This was never about trying harder

Nobody sets out to write "could be more critical" forty times. Faculty already know what good feedback looks like, and they have already put in the effort — that's what the sixty submissions were. The comments get thinner because you're starting from an empty box every single time, on top of a full teaching week, and because writing is slower than judging.

So the fix isn't more effort. It's not starting from nothing.

Your criteria don't change between submissions. Your achievement levels don't change. That means the shape of the comment for weak evidence, or for a strong-but-unfocused argument, doesn't change either — only the details do. Write those shapes down once and grading becomes what it should be: you decide the level, you pick the comment that fits, and you spend your remaining energy on the sentence that's specific to this student.

What we built

That's exactly what Quindaria's rubrics now do. When you build a rubric, you can ask for feedback drafts along with it: for every criterion, at every achievement level, a comment written to the student — naming what the work does, tied to that criterion, ending in a next step.

They're drafts, not verdicts. You mark the level, copy the comment, and make it yours — add the student's name, the paragraph you're thinking of, the thing only you noticed. The blank box is gone; the judgement is still yours.

And because rubrics get shared with students so they know what's being marked, the drafts are kept strictly apart — their own table, their own file, marked for your eyes. Your rubric stays exactly as shareable as it was.

Your course. Your style. Always aligned.

You've already done the difficult part. Before the next stack lands, finish it: write down what you'd say about weak evidence at each level of your rubric, and let the fortieth student get the same quality of comment as the first.