+ Concise, well structured, and best at retaining the core claim.
- Some character counts are wrong, and shorter versions omit continuous use.
Produces the message at 15 characters, 30, one sentence, and one paragraph, noting what each cut sacrifices.
| Category | Marketing › Ad copy |
|---|---|
| Tags | RewritingReformattingMarketer |
Write this message at four lengths. Produce: 1. Ultra-short — under 15 characters. Headline or button. 2. Short — under 30 characters. 3. One sentence. 4. One paragraph, three sentences maximum. For each: - The copy - Character count - *What was sacrificed relative to the longer version* - Where this length is appropriate and where it is not Rules: - Each version must stand on its own. Do not write a short version that only makes sense if you have read the long one. - *Do not truncate. Rewrite.* A cut sentence is not a short version. - Keep the same claim at every length. If the short version has to overstate to fit, say so — that is a signal the claim is too complex for that placement. - No exclamation marks.
Every platform has a different limit and truncating breaks meaning. This rewrites per length and states what the short versions lose, so you can match length to placement.
ChatGPT leads in concision and format compliance. Claude is careful but breaks a core rule, while Gemini is specific yet overly speculative.
+ Concise, well structured, and best at retaining the core claim.
- Some character counts are wrong, and shorter versions omit continuous use.
+ Carefully avoids unsupported claims and maps each length to specific placements.
- The ultra-short copy drops the call-quality claim, violating claim consistency.
+ Actively incorporates the commuter audience and usage context.
- Adds unsupported claims such as guaranteed all-commute performance.
| Criterion | ChatGPT | Claude | Gemini | Leader |
|---|---|---|---|---|
| Instruction following | 8 | 6 | 6 | ChatGPT +33% |
| Accuracy | 7 | 9 | 5 | Claude +29% |
| Specificity | 8 | 9 | 8 | Claude +13% |
| Structure | 9 | 9 | 8 | Tie |
| Right length | 9 | 7 | 6 | ChatGPT +29% |
Scored 1–10 by gpt-5.6-sol with model names hidden (2026-09-24). This is an AI review, not a measurement.
We gave three models the same input and copied their answers unedited. Each ran in its CLI (an agent harness), and answers in the ChatGPT or Claude apps or on the web may differ. Outputs are in Korean.
Message: 3만원대인데 4시간 연속 노이즈캔슬링, 통화음질도 또렷합니다 Product: 무선이어폰 '사운드버즈 X3' Audience: 지하철로 출퇴근하는 20~30대 직장인
※ 글자 수는 공백·문장부호 포함 기준입니다.