Character encoding is declared

Without a declared charset, browsers fall back to heuristics — usually Windows-1252 or Latin-1 on legacy configs — and render multi-byte UTF-8 content as mojibake. Smart quotes, em dashes, emoji, and non-English characters turn into garbage sequences like ’ or é. The Google crawler follows the same fallback and indexes the corrupted text, which then shows up in search results. This is also a minor XSS signal: ambiguous encoding lets attackers smuggle script tags through characters the browser misinterprets.

Most frameworks emit <meta charset="utf-8"> by default — Next.js, Remix, Astro, and Vite-based apps all do. If you are on a custom HTML template or Express-rendered view, add the tag as the first element inside <head> so the browser switches before parsing any content.

// app/layout.tsx or equivalent root template
<head>
  <meta charSet="utf-8" />
</head>

Verify via curl -I that responses also include Content-Type: text/html; charset=utf-8.