# Anatomy of a blog post A fixture showing every block a RomeoSEO post can use: cover image, body copy, figures, a comparison table, lists, a callout, code, and a word from Romeo. - Published: 2026-08-13 - Author: RomeoSEO - Canonical: https://romeoseo.com/blog/anatomy-of-a-blog-post --- This page is the kit, not an argument. Every block a RomeoSEO post can use is on it, named as it appears, so the next article is a copy with the fixture pieces pulled out. The cover above is frontmatter. The title, date, author and reading time are also frontmatter. Everything from here down is the MDX body. ## Body copy Write in US English. Keep the product claims in the prose and the jokes in a bubble. The company is unserious; the product is serious. That split is the whole voice. A post should still work if you delete Romeo's line. If deleting it loses information, it was never a Romeo line. It was product copy wearing a costume. Internal links are ordinary Markdown. They should say where they go, the way [the landing page](/) explains the product and [pricing](/#pricing) states the number. ## An image in the flow Use `
` with real width and height. Markdown `![alt](src)` will fail the build on purpose: an undimensioned image shifts the page, and we sell a product that treats layout shift as a finding.
No stock photography. No generated mascot art. Reuse a screenshot, a Romeo pose, or a file you added under `public/blog/your-slug/`. ## A comparison table GFM tables. First column is the axis; the rest are the things being compared. Keep cells short enough to read on a phone, where the table scrolls sideways rather than shrinking into mush. | | Rank trackers | Article farms | RomeoSEO | |---|---|---|---| | Job | Measure the SERP | Fill a calendar | Ship the fix | | Output | A report | A draft you still have to place | A pull request you approve | | Existing pages | Tracked | Ignored | The default target | | New pages | Out of scope | The whole product | Only when research says a page cannot absorb the query | | Publishes itself | No | Often | Never | That table is a fixture, not a teardown. We are not an Ahrefs replacement and we are not a crawler. We are the layer that turns the data you already have into a change on a branch. ## Lists Some things come in lists. Use bullets when the order does not matter: - Cover image, from frontmatter, with alt text and pixel dimensions - Body copy at a readable measure - Inline figures, comparison tables, callouts - One Romeo line, visually separated, optional Use numbers when the order is the point: 1. Copy `_template.mdx` to `content/blog/your-slug.mdx` 2. Add images under `public/blog/your-slug/` 3. Fill title, description, date, cover; set `draft: false` 4. Open a pull request. Merge publishes. ## A callout, a quote, and a cat > **Constraint** > > Drafts are omitted from production: the index, the sitemap, and `llms.txt`. They render in local development so you can read them before the PR. A blockquote is for a sentence worth pulling, not for decorating a paragraph you already wrote: > RomeoSEO reviews the rankings and content you already have, decides what is actually worth fixing, and ships the fix as a pull request you approve. > Romeo says: I do not write these. I merely refuse to publish them. Inline code looks like `draft: true`. A short block is for something someone might copy: ```yaml title: Anatomy of a blog post publishedAt: 2026-08-13 draft: false ``` ## How a new post actually ships There is no CMS login. The file is the CMS. An agent with access to this repo adds one MDX file, opens a PR, and the next production build puts the post on [the blog](/blog), in `/sitemap.xml`, and in `/llms.txt`. Crawlers that prefer Markdown can fetch `/blog/anatomy-of-a-blog-post/raw`. That is the same loop the product sells. The fixture exists so the next post does not have to invent the blocks. Join the waitlist: https://romeoseo.com