Somebody builds a landing page for a campaign. It goes live, ads point at it, and a decision gets made about whether search engines should see it — usually in the first ten minutes, usually by habit, and almost never revisited.
Both defaults are defensible. Neither is right for every page, and the cost of the wrong one compounds quietly, because an unindexed page produces no evidence that it could have ranked.
What does noindex actually do to a paid page?
It removes the page from organic results and changes nothing about the ads. Search crawlers read the directive and drop the URL from the index; the page still loads, still converts, and still receives every click the campaign pays for.
That separation is the reason the practice exists at all. Ad delivery does not depend on a page being in the search index, so a marketer can hide a page from organic search without touching campaign performance. The bill arrives somewhere else: any organic demand that page could have captured now goes to a competitor, permanently and invisibly.
When should a paid landing page stay out of the index?
When indexing it would put a worse version of an existing page into search results, or would expose something time-bound to people who arrive months later.
Four cases are clear enough to treat as rules:
- Test variants. Three near-identical pages built to test a headline should not all be candidates for the same query. Index one or none.
- Pages that duplicate a service page. If the paid page is your service page with the navigation stripped out, indexing it asks search engines to choose between two versions of your own content.
- Offer pages with an expiry. A promotion that ends in November should not be earning organic clicks in March.
- Pages with parameters in the URL. Campaign-tagged variants multiply quickly and none of them is the canonical version of anything.
When does hiding the page cost you?
When the page is the only thing you have built on that subject. Paid landing pages are frequently the best content a company owns, because somebody was paid to make them persuasive and the results were measured.
Hiding one means the topic has no organic representation, links earned by the page build page authority that leads nowhere useful, and the campaign has to keep paying for every visit indefinitely. A page that ranks continues working after the budget stops. A hidden page does not.
There is a structural version of the same problem. A paid page kept out of the index is also usually kept out of the navigation, which makes it an orphan page in every sense. That is fine when it is deliberate and quietly damaging when somebody later expects it to rank.
Ask what the page is: a variant, or an asset. A variant exists to be compared against another variant and should not be indexed. An asset is the best thing your site says about a subject, and hiding it from search means paying for traffic you could have earned.
Which way should each page go?
Most libraries of paid pages sort cleanly once the question is asked one page at a time.
| Page | Index it? | Reason |
|---|---|---|
| Only page covering this service | Yes | No organic alternative exists |
| Stripped-down copy of a service page | No | Competes with your own better page |
| Headline or layout test variant | No | Multiple versions of one page |
| Seasonal or dated offer | No | Wrong answer for a later searcher |
| Deep guide built for a campaign | Yes | Earns links and rankings after the budget ends |
| Gated or thank-you page | No | Nothing useful to a searcher |
Why is robots.txt the wrong tool for this?
Because blocking a URL prevents crawlers from reading the very directive that would remove it. A noindex tag sits in the page; a crawler that is not allowed to fetch the page never sees it.
The result is the failure most people do not expect: a blocked URL can still appear in results, listed without a description, because the search engine knows the address from links pointing at it but has never been permitted to look inside. The page you were trying to hide is in the index, showing nothing, and the tag that would have removed it is unreachable.
Using robots.txt to keep a landing page out of search. Blocking crawling and blocking indexing are different controls, and using the first to attempt the second produces the worst available outcome — an indexed URL with no content behind it.
What about duplication between the ad page and the service page?
Decide which page you want to rank, then use the control that matches your intent. Noindex removes the paid page from search entirely. A canonical pointing at the service page tells search engines the two are versions of one thing and consolidates the signals onto the version you named.
The canonical is generally the better instrument when the paid page genuinely duplicates an indexed page, because it preserves rather than discards. Noindex is the better instrument when the paid page should not exist in search under any circumstances. Applying both to one page is a contradiction — you are asking for the page to be dropped and for its signals to be passed to somewhere else, and you should not be surprised by which instruction wins.
The same applies to on-page SEO effort spent on a page that will never be indexed. Title tags, headings and schema on a noindexed page do nothing for search, and the hours are better spent on the page you actually want to rank.
What should you check this week?
Pull the list of every URL your campaigns point at and put an index directive next to each one. The audit usually surfaces two things: pages hidden years ago for reasons nobody remembers, and pages nobody meant to index that quietly compete with the service pages.
Then pick the one page in that list that is the best thing your site says about its subject, and decide whether you are comfortable continuing to rent its traffic.
Wondering how many of your ad pages are hidden from search?
We will audit your paid landing pages against your organic footprint and show you which ones are earning nothing because they were told not to.
Get in Touch →Does this apply to your site?
Reading about it is one thing. Point the scan at your own site and see whether this applies to you, and what it is worth fixing.
Free and unlimited. No account, no card, and you get every finding rather than a teaser.