Last-Modified header

Long definition

The header format is fixed: Last-Modified: Wed, 21 Oct 2024 07:28:00 GMT. RFC 7231 mandates GMT and the IMF-fixdate format; deviations break parsers.

The conditional flow mirrors ETag's:

1. First request: server sends 200 OK with Last-Modified: <date>.
2. Client caches the response and stores the date.
3. Subsequent request: client sends If-Modified-Since: <date>.
4. Server compares: file unchanged → 304 Not Modified (empty body). File changed → 200 OK with new content and new Last-Modified.

For SEO, the impact is on crawl efficiency. Googlebot uses Last-Modified two ways.

Reducing crawl spend. A 304 response is a fraction of the bandwidth of a 200. On large sites with many static or rarely-changed pages, this lets Google crawl more URLs within the same capacity ceiling.

Re-crawl scheduling. Googlebot factors the Last-Modified date into how often it returns to a URL. A page that hasn't changed in two years gets crawled less often than one updated weekly. This is also why XML sitemaps use the <lastmod> element — same signal at the sitemap level.

Common pitfalls:

• Lying about freshness. Some CMSes update Last-Modified to "now" on every request to discourage caching. Result: Googlebot recrawls everything, wastes budget, and trusts the signal less.
• CMS template changes invalidating Last-Modified for the whole site. A header/footer edit shouldn't bump every page's modification date. Many platforms get this wrong.
• Resolution mismatch. Last-Modified has 1-second resolution. For high-traffic pages updated multiple times per second, ETag is the correct choice. Most pages don't have this problem.

Last-Modified vs ETag:

• Last-Modified is simpler to implement (a stat() call) and the date carries semantic meaning.
• ETag is more precise and doesn't depend on filesystem timestamps.
• They coexist; clients prefer ETag when both are present, falling back to Last-Modified.

Common misconceptions

• "Updating Last-Modified forces a re-crawl." It signals freshness, but Googlebot decides scheduling. Lying about freshness teaches Google to ignore the signal.
• "Last-Modified must match <lastmod> in the sitemap." They should be roughly consistent — a 30-day delta on a page where the sitemap claims yesterday's update suggests one of the two is broken.
• "Static sites don't need it." They benefit the most. Static-site generators set Last-Modified naturally; ensure your CDN or web server passes it through.