How to give feedback on an essay that students actually act on

We have all written it: a tick, a wavy underline, and "be more analytical" in the margin. It feels like feedback. It took time to write. And it changes almost nothing in the next draft — because the student reads it, nods, and has no idea what to actually do differently on Monday. If you want to give feedback on an essay that students actually act on, the problem isn't effort or care. It's that most feedback tells a student how well they did instead of showing them the next move. This is a practical guide to closing that gap — and an honest look at where a tool can carry some of the load.

Why most essay feedback doesn't land

There are three quiet reasons good teachers write feedback that goes nowhere. The first is that it's a grade in disguise. "Solid work", "developing", "not quite there" — these are judgements, not instructions. They tell the student where they sit; they don't tell them how to move.

The second is that it's unlocated. "Your analysis needs work" floats above the essay without landing on a sentence. The student agrees in principle and then cannot find the paragraph you mean. Feedback that isn't attached to a specific line is feedback a fifteen-year-old cannot use.

The third is that there's simply too much of it. A page bleeding red asks the student to fix twenty things at once, so they fix none. The most useful feedback is often the most ruthless about what to leave out.

Anchor the feedback to the criteria

The single biggest lift in usefulness comes from tying every comment to the criterion it belongs to. A rubric already names the qualities the task is testing — sustains an argument, analyses rather than describes, selects and integrates evidence. When your comment names the criterion out loud, the student stops reading it as a verdict on them and starts reading it as a note on a fixable skill.

Compare "this needs more depth" with: "For the analysis criterion — you quote the metaphor in paragraph three but then retell what happens instead of explaining what the metaphor does. That's the line between the band you're in and the next one." The second version is longer, but it is the only one the student can act on, because it names the criterion, points at the line, and shows the move. It's the same discipline students benefit from when they learn to read a marking guideline like a marker.

A shape for a comment that gets used

You don't need a rigid formula, but a loose shape keeps comments honest and short. A useful comment does three things:

• Names the criterion."On evidence…", "For structure…" — so the student knows which skill this is about, and can see it on the rubric.
• Points at the exact line.Quote or reference the sentence, don't gesture at "your body paragraphs". Located feedback is actionable feedback.
• Shows the next move.Not "improve this", but the concrete thing to try: "after the quote, add one sentence on the effect on the reader before you move on."

Then stop. Two or three comments in that shape will do more than fifteen ticks and a summary sentence. Pick the criterion that is holding the grade down most, and spend your words there.

Feedback the whole class can use

Here is the part that eats your evening: when eight students make more or less the same mistake, you write a variation of the same paragraph eight times. It's worth naming that pattern rather than powering through it. If a criterion is weak across the set, that's a teachingpoint, not a marking one — one strong whole-class example ("here's a paragraph that analyses, and here's one that describes; spot the difference") often lands harder than thirty individual notes saying the same thing. Save your individual comments for what is genuinely individual: the borderline script, the unusual argument, the student you know is capable of more.

Where AI fits — honestly

This is the seam where a marking tool actually helps, and it's worth being precise about what it does and doesn't do. Marktakes the rubric you upload and the essay, and returns a first-pass mark for each criterion with the quoted line it used as evidence — in about a minute per script. That quoted evidence is the useful bit for feedback: it's already located the sentence that shows the analysis, or its absence, so you're not hunting for the paragraph you half-remember. It does not write or rewrite anything for the student, and it doesn't have the final word — every score is yours to confirm or overrule, and you stay the marker of record.

Used that way, the tool changes marking from generating judgements on a blank page to reviewing proposed ones, which frees your time and attention for the comments that need a teacher — the sentence that shows the student their next move. The consistency is a bonus: the same rubric is applied to script one and script thirty, rather than drifting as you tire.

A word on integrity, because students watch what we model. Using AI to help mark — where the teacher confirms every result — is a world away from a student using unapproved AI to produce assessable work, which in NSW is treated by NESA as malpractice. If your students run their own drafts through an AI marker to prepare, that is feedback, not submission — and it's worth teaching them to acknowledge any AI feedback honestly so the line stays clear.

The short version

Feedback students act on is anchored to a criterion, pointed at a line, and pointed toward the next move — and there's less of it than you think you owe. The bottleneck isn't your judgement; it's the time swallowed by locating evidence and re-typing the same note across a class. That is exactly the part worth sharing with a tool that proposes a mark per criterion, with the quoted line, for you to confirm — so your words go where they change the next draft. Teachers get 25 free credits to try it on a real set; see how it handles your rubric before you trust it with a stack.