Revision decisions
Keep a revision log: turn feedback into bounded experiments with a predicted effect, compare versions, and record why you kept or rejected each change.
The idea
A revision experiment names the change, predicted effect and comparison method. It can fail or produce mixed results. Keep enough version history to restore alternatives and explain why you kept or rejected them.
Consent and privacy matter when sharing drafts. Agree the audience, requested scope and use of tools. Do not submit another writer’s manuscript to an AI service without permission. A solo route can use supplied samples and delayed self-review; label it honestly rather than inventing external feedback.
Un ejemplo resuelto
“Test cutting the final explanation, then check whether the new photograph’s significance remains legible” is bounded. “Rewrite until everyone likes it” has no workable stopping condition.
Revision as experiment
A revision experiment names the change, the effect you predict and how you’ll compare. It can fail, or give mixed results, and that’s useful. A revision log records each experiment and why you kept or rejected it, so you can restore an alternative or explain a choice to an editor later.
Design one experiment you expect to reject. If it surprises you, you learned something; if not, you can decline the suggestion with evidence.
Consent and privacy
When you share drafts, agree on who reads them, what you’re asking and how they’ll be stored or deleted. A beta round on Author Beta keeps readers’ copies watermarked and lets you end access when the round closes.
Practise it
Create two experiments from feedback, including one you expect to reject. Record what evidence would change your mind.
For your own project: Apply it to your own draft: feedback matrix and accept/test/decline decisions.
Common pitfalls
- Making changes with no record of why.
- Keeping only the latest version.
- Sharing drafts without agreeing on scope.
Think it through
Can declining a suggestion demonstrate thoughtful revision?
Compare with a suggested answer
Yes, when the underlying issue is considered and the decision is supported by purpose and evidence rather than automatic defensiveness.
Other answers can be just as good if you can point to evidence in the text.
Your turn
Work here or on paper. What you type stays in this browser until you save it to your Desk.
AI assistance optional, off
Off unless you switch it on. Ask about the lesson, get a hint, or have your exercise reviewed against this lesson’s checklist. It never writes the exercise for you. Questions and hints send only your question (we add the lesson). A review sends only the answers you copy into the box. Nothing from your Writing Desk or your books is ever sent, and every request shows you exactly what will leave, and to which AI service, before you send it. Don’t include anything private or confidential. Feedback is advice and may be wrong; it never changes your progress, and every lesson can be finished without AI.
Sign in to use AI assistance.
Try this first
If you can’t predict an effect, the change may be too large. Split it into a smaller experiment.
Still stuck?
Check your work
These are prompts, not rules. Done when: Your work saved to your Desk, or marked done if you did it elsewhere. Self-reviewed, not graded. Counts toward your certificate of completion.
If you get stuck
If you can’t predict an effect, the change may be too large. Split it into a smaller experiment.
Everything you save here is searchable in your Writing Desk and ready to move into a chapter.
Module challenge: Interpret feedback
Three readers dislike a passage, but one read a prior version and another was asked only to check a date. How should the feedback be recorded?
What a strong answer does
Separate version, task, location, actual response and expertise. Do not count the comments as a simple vote on the same question. Seek evidence and a relevant test.
Look for:
Related: Conflicting responses · Read-through audit · All of Revision and Self-Editing Studio
Sources and further reading
- Receiving feedback, The Open University, OpenLearn (Start writing fiction, 7.3.2). Checked Nov 19, 2026. Taking in feedback and deciding what to use.
- Giving feedback to other writers, The Open University, OpenLearn (Start writing fiction, 7.3.1). Checked Nov 19, 2026. Specific, analytical feedback that balances praise and criticism.
This lesson and its examples are original to Author Academy. Sources are cited for ideas and further reading; nothing is copied from them, and no institution endorses this course. Version 1.