How do I cut word count without losing meaning?

Cut structurally before you cut words. Delete the first paragraph, delete the last, remove sentences that restate the one before them, and cut any section that exists only because the outline had four points. That reaches roughly a fifth. Trimming adjectives, which is what most people try first, achieves about three percent and makes the prose worse.

What should I cut from my writing first?

The opening paragraph. Most drafts begin with a runway the writer needed to get started and the reader does not need — general context, a definition, a statement that the topic matters. Delete it and read from the second paragraph. This works on more than half of first drafts, and often you need to delete two or three.

How much should I cut from a first draft?

About twenty percent is typical, and it comes out in chunks rather than word by word. Set a specific target before you start, because cut this from 2,400 to 1,900 is a solvable problem while tighten this up is not, and the vague version produces the adjective-shaving approach that achieves nothing.

Why is cutting so hard?

Because drafts acquire value from the effort they cost rather than from what they contain, so deleting feels like waste. The practical fix is to move rather than delete: keep a scratch file at the bottom of the document. You will almost never retrieve anything from it, but knowing it exists removes most of the resistance.

Does cutting make writing better?

Usually, and twice over: the weak paragraph is gone, and everything around it is now closer together. Density is most of what makes writing feel confident. It also makes every later pass cheaper, which is why cutting should come before rhythm work rather than after.

How to Cut 20% of Your Draft Without Losing Anything

Q: How do I know when to stop cutting?

When removing one more thing would lose information. Signs you have gone too far: the reader now has to infer connections you used to state, examples have been reduced to assertions, and the prose has become a list of claims. Compression past the point of clarity is a different failure, not a better one.

Cutting is the highest-return edit available and the one most writers avoid. Where the fat hides, in what order to remove it, and when to stop.

Almost every draft is about a fifth too long, and the fifth that needs to go is rarely where the writer thinks it is.

People trying to shorten a piece tend to trim adjectives and shave clauses — the slow, painful method that removes maybe three percent and makes the prose slightly worse. The material actually available for removal is structural, and it comes out in chunks.

This is where it hides, in the order worth attacking it.

Why cutting is the highest-return edit

Removing a weak paragraph improves a piece twice: the paragraph is gone, and everything around it is now closer together. Density is most of what makes writing feel confident.

There is also an arithmetic reason to cut before anything else. Every other editing pass — rhythm, word choice, fact-checking — costs time proportional to length. Cutting first makes all of them cheaper, and it stops you polishing sentences you were going to delete. Doing it in the other order is one of the most common ways writers waste an afternoon.

The obstacle is not skill. It is that cutting means discarding work you already did, and drafts acquire a value from the effort they cost rather than from what they contain.

1. The runway

Delete the first paragraph. Read from the second.

Most drafts open with a paragraph the writer needed in order to start — general context, a definition, a statement that the topic is important. It is scaffolding, and it comes down once the building stands. Frequently the first two or three paragraphs go.

Do the same at every section heading. Sections have their own miniature runways: "Now that we've covered X, let's look at Y." That sentence is doing navigation the heading already did.

Typical yield: 5–8%.

2. The landing

Then delete the last paragraph.

Summary conclusions are the most reliably deletable thing in any draft. A reader who has just read your article does not need it recapped, and a conclusion that only restates is dead weight in the position of maximum emphasis.

Keep an ending only if it does something the body did not: an implication, a next step, a limitation you have been holding back, a genuinely unresolved question. "In conclusion, sentence variety is essential for engaging writing" is a mirror. Cut it and end on your last real point.

Typical yield: 3–5%.

3. The restated sentence

Now go paragraph by paragraph and read only the last sentence of each.

Models and tired writers both do the same thing: state a claim, then restate it in different words to round the paragraph off. It feels like reinforcement. It reads as padding.

Regular maintenance extends equipment lifespan considerably. Machines that are serviced on schedule simply last longer than those that are not.

The second sentence is the first sentence wearing a hat. Delete it.

This one is mechanical and fast — you are scanning for paragraphs whose final sentence contains no new information.

Typical yield: 4–6%.

4. The hedge and the filler opener

Search for these specifically:

"It is important to note that"
"It's worth mentioning"
"In today's" (anything)
"When it comes to"
"The fact that"
"In order to" → "to"
"There is / there are" at the start of a sentence
"Considering these factors" and similar empty participial openers

Nearly all of these delete cleanly, and the sentence improves. "It is important to note that maintenance matters" becomes "Maintenance matters," which is both shorter and more confident.

Also hunt intensifiers — "very," "really," "significantly," "substantially," "incredibly." Each one either gets replaced with a number or gets cut. "Improved significantly" tells the reader nothing; "cut review time from two days to three hours" tells them everything and costs a similar number of words.

Typical yield: 3–5%.

5. The section that is there for symmetry

The biggest single cut, and the hardest to see.

Outlines produce sections. Sometimes a section exists because the outline had four points and this was the fourth, not because you had anything to say about it. It will be the shortest, vaguest section in the piece, and it will be padded to look the same size as the others.

Read your headings on their own. For each, ask what it says that the reader could not have guessed. If the answer is nothing, delete the whole section — do not try to improve it. A three-section article where every section earns its place beats a four-section one with a passenger.

Uneven section lengths are a feature. Real writing gives disproportionate space to the interesting part.

Typical yield: 0–15%, depending on the draft.

6. The example that proves the same thing twice

If you have given two examples of one point, one of them goes. Keep the more specific.

Writers double up on examples out of anxiety that the first one did not land. It almost always did. The second example is insurance, and insurance is expensive in a reader's attention.

The exception: two examples that demonstrate genuinely different aspects of the point, or a contrasting pair where the contrast is the point. Those earn their place. Two illustrations of the same idea do not.

Typical yield: 2–4%.

The order matters

Structural cuts before sentence cuts, always. There is no point tightening the clauses in a paragraph you are about to delete, and every minute spent doing so makes you more reluctant to delete it — the effort you have invested becomes an argument for keeping it.

So: runway, landing, restated sentences, filler phrases, redundant sections, doubled examples. Big to small. Sentence-level polishing comes after, on a draft that has already lost its fifth.

The techniques that make it possible

Cut into a holding file. Do not delete — move. A scratch document at the bottom of the draft, or a separate file. Knowing the material still exists removes almost all the resistance, and you will essentially never retrieve anything from it. The point is psychological, not practical.

Set a target before you start. "Cut this from 2,400 to 1,900" is a solvable problem. "Tighten this up" is not, and produces the adjective-shaving approach that achieves nothing.

Cut on a different day. You cannot see your own scaffolding while the reasoning that produced it is still live in your head. Twenty-four hours is usually enough for a draft to become someone else's.

Read it aloud. Padding is audible. Where you get bored reading your own sentence, a reader stopped two sentences ago.

Cut your favourite thing. If a paragraph is clearly the best-written in the piece but does not advance the argument, it is a passenger with good manners. This is the hardest cut and often the most improving one.

When to stop

Cutting has a floor, and going past it does damage.

Signs you are done: removing anything else would take out information, the piece reads faster than it did without feeling rushed, and every section earns its heading.

Signs you have gone too far: the reader now has to infer connections you used to state, examples have been reduced to assertions, and the prose has become a list of claims. Compression past the point of clarity is not concision — it is just a different failure.

A useful stopping rule: cut until removing one more thing would lose information, then stop. Not until you hit a word count, and not until it feels lean, which is a feeling that keeps moving.

What cutting will not do

It will not give a piece a point it never had. If the draft surveys a topic without claiming anything, a shorter version is a shorter survey. That is a structural problem and the fix is upstream — see how to write an opening that earns the next paragraph for the promise a piece has to make, and the editing checklist for the passes either side of this one.

It also will not add the specific detail that makes writing worth reading. Cutting removes what is not needed; it cannot supply what is missing. Most drafts need both, and the additions usually come from you rather than from any tool — a number from your own work, an incident, an opinion you would defend.

Our Text Humanizer does a version of steps 3 and 4 automatically, cutting hedging and restated phrasing while holding the meaning fixed. Steps 1, 2, and 5 need judgement about what your piece is for, which is not something a rewriting tool can have.

Cutting is one pass of several. The full sequence, from checking facts to reading aloud, is in how to write like a human.