Comment-language mining

댓글 언어 채굴

Turn recurring customer phrases in comments into copy candidates to verify.

···
html
<div class="comment-mining"><div class="comments"><span>“I can't find Monday's note.”</span><span>“Where did that note go?”</span><span>“I keep losing the next step.”</span></div><div class="extract"><small>RECURRING LANGUAGE</small><b>Find the note you need.</b><i>3 sample comments → one headline candidate</i></div></div>
css
.comment-mining{display:flex;align-items:center;gap:clamp(8px,2vmin,20px);width:89%;max-width:700px;font-family:sans-serif}.comments,.extract{min-width:0;flex:1}.comments span{display:block;margin:clamp(4px,1.2vmin,10px) 0;padding:clamp(7px,1.6vmin,13px);border:1px solid var(--line);border-radius:10px;background:var(--surface);font:750 clamp(10px,2.1vmin,16px)/1.1 sans-serif;animation:scan 6s infinite}.comments span:nth-child(2){animation-delay:1s}.comments span:nth-child(3){animation-delay:2s}.extract{box-sizing:border-box;padding:clamp(9px,2.4vmin,20px);border-left:5px solid var(--accent);background:var(--surface);animation:extract 6s infinite}.extract small{display:block;color:var(--accent);font:900 clamp(10px,1.6vmin,12px) sans-serif}.extract b{display:block;margin:.5em 0;font:900 clamp(12px,3vmin,25px)/1.1 sans-serif}.extract i{font:700 clamp(10px,1.7vmin,13px) sans-serif;font-style:normal;color:var(--muted)}@keyframes scan{0%,17%{border-color:var(--accent);transform:translateX(4px)}30%,100%{border-color:var(--line);transform:none}}@keyframes extract{0%,12%{opacity:.4;transform:translateX(-6px)}25%,86%{opacity:1;transform:none}100%{opacity:.4;transform:translateX(-6px)}}@media(min-width:700px){body > *{transform:scale(1.4)}.lines p,.branches div,.cards div{cursor:pointer}}@media(min-width:700px){
  .creator-scene small,.creator-scene .meter,
  .journey .track,.journey .panels small,.journey .panels p,
  .disclosure .ad small,.disclosure .labels span,
  .comment-mining .comments span,.comment-mining .extract small,.comment-mining .extract i,
  .inbox .inbox-top,.inbox .inbox-note,.inbox .sender,.inbox .preview,
  .hierarchy small,.hierarchy span,
  .tone .constant,.tone .tone-track,.tone .tone-messages small,.tone .tone-messages p,
  .naming .ylabel,.naming .xlabel,.naming .point,.naming .choice{font-size:12px}
}

Comment-language mining collects the words customers use voluntarily instead of relying only on a brand's vocabulary. TikTok's 2026 report treats comments as part of the creative canvas and recommends listening to language in ongoing conversations.

The demo gathers a repeated 'can't find it' phrase from fictional comments and turns it into a headline candidate. Frequency does not make a comment representative of all customers. Preserve context and avoid exaggeration.

When to use

Compare recurring public comments with interview language before writing a new headline.

Open as page ↗