Told "too wordy" three times,
we measured 65 popular posts

1Edited by feel, three times

The first draft was a compressed version that kept only the key points. It got the note "the content is too compressed."

The second put in real usage examples and a conversational voice. This time it was "lower the overall tone." The sentences that pronounced things like aphorisms were what snagged.

The third we rewrote in a tone of observing and suggesting. The note that came back was "the sentences are wordy."

All three times we edited earnestly, and all three times it was by feel. We stopped here.

2We decided to measure

We decided to bring the standard from outside. Writing that actually gets read and circulated on social media — we gathered the full text of 65 popular posts.

We measured three things.

• Sentence length. How many characters is one sentence (median).
• Density per line. How many sentences fit in one visual line.
• Density per paragraph. How many sentences are bunched into one paragraph.

We measured our own card drafts by the same yardstick.

3A surprising result — our sentences were shorter

The median sentence length of the real corpus was 32 characters. Our cards were 27.

Meaning the "wordy" writing had shorter sentences than the writing that gets read.

The difference was in density.

The real posts break one sentence per line and give it room to breathe. 83% of lines held 1–2 sentences.

We had bunched 4–8 sentences into one paragraph. 69% of paragraphs had four or more sentences.

Even sentences of the same length read like a wall once bunched together.

4The three true faces of wordiness

Setting the numbers and the originals side by side, the causes settled into three.

• Paragraph density. Just what we saw above. Not broken up, it becomes a wall.
• Stage-setting sentences. Lines like "the scene is always similar" or "what's interesting is what comes next." Writing that gets read has almost none of these — it goes straight to the point.
• Dash-chaining. Splicing sentences together with "— so" or "— this approach" makes for one long breath.

All three were habits born of trying to write well. It was a matter of breath, not of sentences.

5Same content, after editing

The standard we derived is short. One sentence per line. Conclusion first. Delete stage-setting. One or two sentences per item.

We also decided what not to adopt. The emoji and exaggerated hooks common in the real corpus, we don't use. Source labeling we keep even though the real posts lack it. Borrow someone else's rhythm, but what is ours we decide ourselves.

6If you want to try it

This method works even beyond card news. Whether it's a newsletter or an announcement, the order is the same.

1. Gather the reference material. 20–50 pieces in your field that actually get read. Not the ones that look good, but the ones with confirmed engagement.
2. Measure only three things. Median sentence length, sentences per line, sentences per paragraph. You can have AI do the counting.
3. Measure your own writing by the same yardstick. Instead of impressions, you get two lines of numbers.
4. Fix only where you diverge. In our case we left the sentences as they were and changed only how we broke them up.
5. Write down what you won't adopt. Copy everything and it stops being your writing.

Collecting and measuring, it takes a day. Faster than editing by feel three times.

7The gist

The answer to the "too wordy" feedback wasn't "write shorter." Before measuring we didn't know the cause; after measuring it was done in one go.

Measure instead of sense it — the same prescription worked for writing, too. This very piece you're reading was written by that standard.