Your pages competing with each other

What it looks like

A search where your site appears, but the page shown keeps changing between two candidates from one week to the next.

Or two pages each ranking modestly for the same phrase, where neither reaches the position one strong page would.

Or a page that used to rank slipping just as you published something new on an adjacent subject.

The flickering is the clearest symptom and the easiest to miss, because you have to be watching one query over several weeks to notice it at all.

• The page shown for a search keeps changing week to week
• Two of your pages each place modestly where one should place strongly
• A page slips just as you publish something adjacent
• Two pages with near-identical titles
• Internal links pointing at different pages for one subject
• You cannot say in one sentence who each page is for

Why it costs you

Links, attention and relevance signals that should accumulate on one page are split across two. Each page is weaker than the combination would be.

The engine also has to guess which page you meant, and it may choose the weaker one, or a page you never intended to rank at all.

And your own internal links become inconsistent, because different pages on your site point at different versions of the same subject.

Links, attention and relevance that should accumulate on one page are split across two, and the engine may choose the weaker one.

How it happens

Writing a page per keyword variant rather than per intent, which is the most common cause on small sites.

Publishing a new article on a subject an older page already covered, without checking what already exists.

Location pages that overlap heavily, or service pages where two services are really one described two ways.

Writing a page per keyword variant rather than per intent is the commonest cause on a small site.

Finding it

Search Console, filtered to one query, showing which page appeared. Two or more of your pages alternating is the signature.

You can also search your own site for a phrase and see how many of your pages cover it. Anything above one is worth a look.

Sort your pages by impressions and look for pairs with similar titles. Similar titles usually mean similar intent.

Do this before writing anything new, not after. The cheapest cannibalization to fix is the page you decided not to publish.

Filter the performance report to a single query and look at which page appeared. Two of yours alternating is the signature, and it takes a minute per query.

Merging is usually the answer

Take the stronger page, move anything useful from the weaker one into it, and redirect the weaker one to it permanently.

The merged page inherits the standing of both and answers the subject better than either did, which is often a larger improvement than months of optimisation.

Update your internal links to point at the survivor, so the site stops disagreeing with itself.

The merged page inherits the standing of both and answers better than either did, which is often a larger gain than months of tuning — see thin content.

When to keep both

When the intents genuinely differ. A guide explaining a process and a service page offering it serve different people and can both rank.

In that case make the difference obvious: different titles, different opening answers, and a clear link between them.

The test is whether you can describe in one sentence who each page is for. If you cannot, they are one page.

Where both stay, link them to each other plainly, so a reader who landed on the wrong one can move across in a click.

Where both stay, link them plainly to each other so a reader who landed on the wrong one can move across in a click.

How ours prevents it

Pages here are planned one per intent, and each new page is measured against existing pages before it can publish.

A page too close to one that already exists is refused rather than added, which is the same judgment an engine makes but applied before the page exists.

Every page passes a check that runs before anything publishes, validated against Google Search Essentials and modern web standards.

A page too close to one that already exists is refused rather than added — the same judgment an engine makes, applied before the page exists.