Make a revision plan before changing sentences

Read once to identify the work

Opening a draft with the intention to “make it better” invites you to fix whatever catches your eye first. A repeated adjective may receive immediate attention while a chapter with no clear function survives untouched. Begin with a reading pass whose purpose is diagnosis. Record observations without stopping to solve every sentence.

Describe the problems in terms of their effect. “Chapter four is weak” gives you little direction. “Chapter four repeats the history from chapter two and delays the first worked example” identifies both duplication and a consequence for the reader. In fiction, “I cannot see why the protagonist accepts the offer” is a more useful starting point than “the protagonist needs work”.

Keep evidence with each observation. Note the location, the expectation established earlier and the point where the reading experience changes. This allows you to return to the problem later without rediscovering it. It also helps distinguish a recurring issue from a single awkward passage.

Find the decisions that control other changes

Some revisions alter the conditions for everything below them. Changing the intended audience may affect vocabulary, examples and assumed knowledge throughout a nonfiction book. Giving a different character the central decision may change viewpoint, scene order and the ending of a novel. These decisions need attention before the dependent passages receive detailed polishing.

Draw a simple dependency when necessary: decide what the chapter must explain, choose the supporting example, then revise the transitions. This need not become a project-management exercise. A few arrows or an ordered list can prevent you from spending an afternoon refining a section whose purpose is still undecided.

For an illustrative manuscript about repairing household furniture, the author might discover that beginners need to identify the cause of a loose joint before comparing repair methods. Moving diagnosis earlier changes several chapters. The useful first task is to establish that learning sequence. Rewriting the introduction to each method can follow once the sequence is settled.

Limit the active plan

A complete list of flaws is rarely a workable daily plan. Select a small set of priorities that will most affect the book's meaning, coherence or usability. Keep the remaining observations in a separate backlog so they are preserved without competing for immediate attention.

Give each priority an action and an observable result. “Improve pacing” is broad. “Remove the repeated journey summary and bring the confrontation into the same chapter as the discovery” can be attempted and evaluated. The result may still fail, but you will know which change you made and what it was intended to achieve.

Avoid declaring every issue urgent. A spelling inconsistency and an unsupported central claim do not have equal consequences. Nor do they require the same kind of attention. Group work by the decisions it demands: research, restructuring, scene development, sentence revision and final checking. Use only the groups your manuscript actually needs.

Leave room for discoveries. A structural revision often exposes a missing connection that was hidden by the old arrangement. Update the plan when the evidence changes, while recording why a new task has replaced an earlier priority. Otherwise the plan can become another rigid object you keep revising around.

Work in passes that have a purpose

A pass is useful when it directs your attention toward a coherent question. You might follow one character's decisions through the manuscript, check whether every stage of an explanation has adequate support, or read only the beginnings and endings of chapters to assess progression. A pass described merely as “another edit” provides less guidance.

Try to complete the intended structural change across the relevant span before judging the prose locally. Moving an argument's conclusion may temporarily leave rough transitions. Repair them enough to read the new order, then assess whether the logic works. There is little value in perfect transitions between paragraphs that still belong elsewhere.

Record completed decisions briefly. This helps you avoid reopening them simply because another option occurs to you later. You can reconsider a decision when new evidence warrants it, but a passing preference should not erase the reason for the earlier choice.

Check the result against the original problem

After each major change, return to the observation that prompted it. Can the reader now identify the missing motivation? Does the worked example arrive when the necessary terms are available? Read the surrounding material as well, because a local improvement can create a new inconsistency elsewhere.

When the principal problems have been addressed, use a fresh read to decide the next stage. The manuscript may need another structural pass, a focused outside response or sentence-level work. The plan has succeeded if it has made those choices clearer and directed effort toward changes that matter to the reader.