Most essay feedback arrives after the grade, when students have already stopped caring about the paper. They glance at the score, skim the comments, and move on to the next assignment. The information is useful, but the timing wastes it.

Draft feedback fixes the timing. When a student receives comments while the essay can still be changed, the feedback becomes a set of instructions instead of a verdict. For a novel like The Bell Jar, where the ideas are subtle and the evidence needs careful handling, that cycle of drafting and revising is where the most learning happens.

The obstacle is workload. Reading every draft and writing detailed feedback for a hundred students is not realistic for most teachers, especially when it means reading each essay twice.

This is where AI feedback tools are practical. They can respond to drafts quickly against your rubric, letting students revise before you do your own read of the final version.

Setting Up a Draft Cycle

A simple structure is to have students submit a full draft a week before the final due date. Each student receives rubric-based feedback and writes a short revision plan explaining what they will change. The final paper is then graded by you, with the plan attached so you can see how the student responded.

• Draft due one week before the final essay, with a complete introduction and conclusion
• Feedback tied to rubric criteria so students know what each comment addresses
• A short revision plan naming two specific changes the student will make
• Final submission with a brief note explaining what changed and why
• Teacher grading of the final essay with attention to how well the feedback was used

Making the Feedback Useful

Draft feedback works best when it is limited. Three to five priorities are more effective than fifteen. Ask the tool or yourself to focus on the biggest issues first, such as the thesis, the use of evidence, and the analysis of key passages, and leave sentence-level polishing for later.

Students also need to know what good revision looks like. Show a before-and-after of a paragraph, ideally from an anonymous earlier student, so they can see how a summary paragraph becomes an analytical one.

Keeping the Student's Voice

One concern with AI feedback is that students may rewrite essays to please the tool. The remedy is to make the feedback descriptive, not prescriptive. Comments that point out a gap, such as a claim without evidence, let the student decide how to fix it.

Be clear about academic integrity too. Feedback tools are for improving the student's own writing, and the policy on generating text with AI should be stated in the assignment instructions.

Measuring Whether It Worked

Compare draft scores with final scores across the class. If the average moves up and the weakest papers improve most, the cycle is doing its job. If nothing changes, the feedback may be too vague or the revision step too optional.

Ask students what helped. Their answers will tell you which comments they used and which they ignored, and that is the best guide for adjusting the process next time.