- Oct 31, 2026
- 3 min read
Page 1, Zero Clicks: Giving Astro Articles a Separate SEO Title
The monthly analytics report has a list called page 1, no clicks. The first time I ran it, six articles were on it, at positions 5 to 8, with impressions and zero clicks. The reason was obvious once they were side by side:
| Page heading (written for readers) | What people searched |
|---|---|
| Cloudflare Email Obfuscation Fooled My Deploy Check | cfemail, data-cfemail |
| Mailpit vs Mailtrap vs listmonk: What Each One Is Actually For | mailpit vs mailtrap |
| GitHub Releases and Obtainium for Android APK Updates | obtainium could not get id from apk |
The headings tell a story. On a results page, the person searching
data-cfemail sees a title that doesn’t contain it.
Two titles, two jobs
- The H1 is read by someone already on the page. A story is fine.
- The
<title>is read in a list of ten results. It should start with the thing they typed.
In my site both came from one title field. So, an optional second one.
The change
The content collection schema:
// src/content.config.ts
schema: z.object({
title: z.string().default("Untitled Article"),
// Optional <title>/og:title for search results (≤ 60 chars); the page H1
// and article lists keep `title`.
seoTitle: z.string().optional(),
description: z.string().optional().default(""),
// …
}),
The article page, where the SEO object is built:
const seo = {
title: post.data.seoTitle ?? post.data.title,
description: post.data.description,
image: post.data.image,
};
And in the front matter:
title: "Cloudflare Email Obfuscation Fooled My Deploy Check"
seoTitle: "data-cfemail: Decode Cloudflare-Obfuscated Email Addresses"
description: "What Cloudflare's data-cfemail attribute is, why curl and grep can't see the email behind it, and ten lines of Python that decode it (a one-byte XOR)."
The H1, the article list and the structured data keep title. Only the
<title> and og:title change. URLs don’t change, so nothing to redirect.
Rules I used for the rewrites
- Lead with the searched term, exactly as typed, if the article really answers it.
- ≤ 60 characters, before any site-name suffix.
- Description ≤ 160 characters, saying what the reader gets, not what I did.
- Only for pages already on page 1 or 2. A better title won’t lift a page from position 70; it turns impressions you already have into clicks.
The same idea on a homepage
My engineering site ranks 11th for its own field’s name, and its homepage
title didn’t contain the exact phrase. Three changes, none of them
redesigns: the phrase in the <title>, in the small label above the hero
heading, and as a link in the first paragraph to a new glossary page that
answers “what is …?”. The big design headline stayed.
Did it work?
Too early: Google needs to recrawl and then show the new titles for a few weeks. I’ll check the same list in mid-November and report back.
Takeaways
- H1 for readers,
<title>for the results page. They’re allowed to differ. - An optional field with a fallback keeps every existing article working.
- Retitle what already ranks; don’t expect titles to fix rankings.