Cross-Posting Without Wrecking Yourself: A Canonical Tag Primer

A writer with a blog and a Medium account eventually asks the obvious question: can I just put the same piece in both places? The answer is yes, it is a sound strategy, and roughly half the people who do it accidentally arrange for the copy on somebody else’s domain to outrank their own.

The fix is not complicated. It is one HTML tag and a decision about publishing order, and getting both right takes about thirty seconds per post.

The actual problem

When the same article exists at two addresses, a search engine has to decide which one to show. It will not show both, and it will not ask you.

Left to itself, it picks the one with stronger signals — more links, more history, more authority. Between a large publishing platform and a personal blog nobody has linked to yet, the platform wins essentially every time. Your piece ranks, traffic arrives, and it arrives at a page where the readers are somebody else’s, the design is somebody else’s, and any subscription revenue is subject to somebody else’s terms.

This is not a penalty. Nobody is punishing you for duplicate content; that particular piece of folklore has been wrong for years. It is simply a choice being made on your behalf, badly, because you did not make it yourself.

What a canonical tag does

A canonical tag is a line in a page’s HTML that says: this page is a copy, and the original lives over there. It is a single link element in the page head, with a rel attribute of canonical and an href pointing at the address you consider authoritative — in this case, the copy on your own domain.

Place that on the republished copy, pointing at your version, and you are telling every search engine to credit the original. Ranking signals earned by the copy consolidate onto your page. The copy still gets read by the platform’s own audience, and you still get the readers — you simply stop competing with yourself.

It is a strong hint rather than an absolute instruction, but it is honoured reliably when the two pages genuinely are the same content and the tag points somewhere sensible.

Two failure modes account for nearly every case where it does not work. The first is pointing the canonical at a page that no longer exists, or that redirects somewhere else — a search engine asked to credit a broken address will simply ignore the instruction and fall back to choosing for itself. The second is declaring two pages equivalent when they are not: if the republished version has been trimmed, retitled or substantially reformatted, the claim of equivalence is false and the tag carries correspondingly less weight. Keep the copies genuinely identical and keep the target alive, and the mechanism behaves exactly as documented.

The order matters as much as the tag

Publishing order does work that the tag alone cannot, and this is the part people skip.

Publish on your own site first. Let it be indexed — usually a day or two. Your version now exists first and is discovered first, which is a signal in its own right.

Then republish elsewhere, with the canonical pointing home. The copy arrives second, declares itself a copy, and points at a page the search engine has already seen and understood.

Reverse that order and you are asking a search engine to demote a page it indexed first in favour of one it found later. It will usually comply, but you have made the job harder and taken on risk for no gain.

How each platform handles it

Support varies, and the variation tells you something about each platform’s attitude.

Hashnode handles this well. There is a dedicated field for the canonical URL when you publish, and it is presented as a normal part of the workflow rather than something buried. That is consistent with the rest of the platform’s behaviour, as noted in our Hashnode review.

Medium supports canonical tags through its import tool: give it the URL of your original and it pulls the piece across with the canonical set correctly. Compose the post from scratch in Medium’s editor instead, and you get no canonical and a direct competitor for your own piece. The import route is strictly better and takes less effort.

Dev.to and similar developer communities generally expose a canonical field, often directly in the post’s front matter.

LinkedIn newsletters and most social platforms do not offer one at all. Treat those as excerpt-and-link territory: post the first few hundred words, link to the full piece, and do not republish in full.

Your own platform needs to produce the canonical on the original too, pointing at itself. Every platform in our best blogging platforms guide does this by default, but it is worth confirming in the page source rather than assuming.

Publish where you own the address. Republish where the readers already are.

When not to cross-post at all

Three cases where the whole strategy is wrong.

When the piece is the reason someone would subscribe. If it is your best work and your goal is a readership on your own site, giving it away on a platform that keeps the reader relationship is working against yourself.

When you would republish everything. A blog that exists in full somewhere else has no reason for anyone to visit it. Cross-post a selection — the pieces aimed at a platform’s particular audience — and keep the rest exclusive.

When the platform rewrites your piece. Some platforms alter formatting enough that the copy is meaningfully different. Canonical tags assume equivalence; if the versions have diverged, you are managing two articles rather than one.

The workflow, compressed

For anyone who wants this as a checklist:

1. Write and publish on your own site.
2. Wait for it to be indexed — a day or two is usually enough.
3. Republish on the second platform using its import tool where one exists.
4. Set the canonical to your original URL. Verify it in the page source; do not take the interface’s word for it.
5. Never republish in full anywhere that cannot set a canonical.
6. Re-check twice a year, because platforms change these fields without telling anyone.

Done in that order, cross-posting is close to free: you keep the search traffic and the authority, and you borrow the other platform’s audience for the price of a few minutes. Done in the wrong order, you have spent months building somebody else’s domain and wondering why your own gets no visitors. The mechanics of that borrowed audience — what it is genuinely worth, and what it is not — are in where readers actually come from.