How Twitter Card tags work
When a link is posted on X, the platform fetches the page and looks for tags with a twitter prefix. The name has outlasted the rebrand, so the markup is still written as twitter:card, twitter:title, twitter:description and twitter:image. The first of those is the one that changes the layout: a summary card puts a small square image beside the text, while the large-image variant gives the picture the full width of the post.
The part people miss is the fallback. If a page has no twitter tags but does have Open Graph tags, X will generally use those instead. That is why a page can preview correctly on X without a single twitter tag in its source — and why the honest answer to “do I need them?” is usually “only where you want X to differ from everywhere else”.
Why Twitter Card tags matter
A post on X is short. The card is often larger on screen than the words above it, so the image and headline are carrying the click, not your copy. Choosing the large-image card for an article and the compact one for a routine page is a real decision about how much attention a link deserves.
They also let you write for a different audience without touching the rest of the site. The wording that suits a professional network is rarely the wording that suits X, and separate tags mean you can be blunter or more specific there while leaving the page’s own title and the previews on other platforms alone.
Common mistakes with Twitter Card tags
The first is duplicating Open Graph exactly. If the title, description and image are identical, the extra tags add weight to the page and nothing to the preview; leave them out and let the fallback do its job.
The second is a broken or unreachable image. The file has to be publicly fetchable, and a large one on a slow connection can time out before the card is built, so a heavy hero photo is a genuine cause of blank previews. The third is expecting to see changes instantly: the platform caches what it fetched, and the old card can persist after the page is fixed.
Finally, X no longer runs the public card validator it once offered, so a workflow built around that tool has quietly stopped working. Checking now means drafting a post with the URL and looking at the preview before you publish it.
How to act on it
Decide first whether X is a channel you genuinely use. If it is not, set good Open Graph tags, let X inherit them and spend the time elsewhere. If it is, set the card type deliberately per page, give articles their own image and a headline written for that feed, and credit the account so the card carries your name.
Then test on a phone. Cards are cropped differently at small sizes and text placed near the edge of an image disappears. Where link previews are part of how you distribute posts, fold the check into your social media marketing process so nothing goes out untested.
