Your competitor copied your archive. The original usually holds.

Someone has republished your posts. Perhaps verbatim, on a domain nobody has heard of. Perhaps run through a rewriting tool, examples swapped, and put out under a competitor's name. The instinct is to treat this as theft in progress and rewrite the affected pages so yours are different again, and that instinct is almost always wrong.

Google picks one page, and the copy starts behind

Duplication is an indexing problem before it is a competitive one. When Google finds pages whose primary content is the same or very similar, it clusters them and picks one to represent the group. Its canonicalization documentation describes this plainly: Google chooses the page that, on the signals its indexing process collected, is the most complete and useful for search users, and marks that one as canonical. The rest are crawled less often and generally do not show.

The decisive words there are the signals its indexing process collected. Those signals were collected before the copy existed. Your page has a crawl history, whatever links it has earned, and the internal links pointing at it from the rest of your archive. The copy has the text and nothing else. It enters a cluster that has already been scored, carrying the one asset that was never scarce.

This is also why duplication by itself is not a punishment. Google states that some duplicate content is normal and not a violation of its spam policies. Deduplication is a display decision — show one version rather than five — not a sanction against either party.

The edit meant to disguise the copy is the part Google names

The more sophisticated version of this, where a competitor runs your archive through a model and publishes the output as their own, feels more dangerous than a straight copy. In ranking terms it is usually worse for them.

Google's spam policies name the behaviour directly under scraping: copying content from other sites, modifying it only slightly — substituting synonyms, or using automated techniques — and republishing it. The scaled content abuse section names the same move again, listing automated transformations such as synonymising and translating among the ways unoriginal pages get produced at volume.

So the rewrite does not launder the copy into something original. It moves it out of the merely duplicated category and into a named policy. The relevant standard is the one we keep coming back to: Google's stated objection is to unoriginal content that provides little value, no matter how it was created. A machine-spun version of your page is the purest example of the thing the policy describes.

The cases where the copy does win are about indexing, not prose

This argument has real exceptions, and pretending otherwise would be the same overconfidence we are warning against.

The copy can win when it is indexed first. If your publishing pipeline is slow, your sitemap stale, or your content assembled client-side in a way that delays crawling, the scraped version can be the one Google sees and clusters around while yours is still waiting. Speed to index matters more here than anything about the writing.

It can win when the site hosting it is far stronger than yours. Canonical selection favours the page judged most complete and useful, and site-level signals feed that judgement. A copy on a large, heavily crawled domain can take the cluster from a small site that published first. Google's own list of canonicalisation factors is short — protocol, redirects, sitemap inclusion, `rel="canonical"` annotations — and most of what decides the outcome is not a dial you control.

And it can win when you syndicated the thing yourself and forgot: a guest post, a partner republication, a wire distribution. Those are not attacks. They are agreements that nobody set a canonical for.

Rewriting what they took concedes the page

The reason rewriting is the wrong move is mechanical. The crawl history, the links, and the internal linking that were winning the cluster are attached to the page as it stands. Rewriting it to be different from the copy throws away the position it already held in order to solve a problem it was already winning.

The productive responses are duller. Get published and crawled quickly, so originals are always indexed first. Link new posts from the archive you already have, which is the cheapest version of the same protection and worth doing anyway. Keep publishing, because a copier can take what exists and cannot take what you publish next.

The legal route exists but is shaped differently from how people imagine it. Google says that when it receives a significant volume of valid copyright removal requests about a site, it can demote other content from that site. That is a pattern instrument. One request about one page is a removal, not a ranking lever, and treating it as one leads to disappointment.

What changes if you accept this

Check before reacting. Search a distinctive sentence from the post in quotation marks and see which URL Google returns. If it is yours, nothing is broken and the copy is costing you nothing but irritation. If it is theirs, look at indexing order and site strength rather than at your prose, because those are the factors that actually decided it.

And if nobody copied you and a competitor is simply ahead, that is a different problem with a different answer. Either way, the hour is better spent on the next post than on rewriting the last one.