What to Noindex on an Affiliate Site: Tag Pages, Author Archives, Pagination, Search Results, and the Pages That Should Stay
WordPress makes a page for everything. Publish one guide with three tags and you've also made three tag archives, a date archive for the month and the year, an author archive, an attachment page per image, and a new entry on every paginated list the post appears in.
On a fifty-guide site that's a few hundred thin pages beside fifty real ones, each one a URL Google can crawl, index and judge the site by. The categories and tags post covers why those archives leak crawl attention on a new site; this is the page-by-page decision and the mechanics of carrying it out without taking down pages that should stay.
This post covers the decision for each page type WordPress generates (noindex, index, or remove), the difference between noindex, robots.txt and a canonical and which does what, how to set each in the common SEO plugins, how to check the result in Search Console, and the pages people noindex by mistake and pay for.
The short answer
Noindex tag archives, date archives, attachment pages, internal search results, and author archives on a one-author site. Keep posts, pages, the homepage and real hubs indexed. Make category archives either the hub (indexed, with text) or noindexed, never a thin duplicate of the hub.
Leave pagination indexed with a self canonical. Use noindex, not robots.txt, to remove pages from results. Then check Search Console's "Excluded by noindex" count a month later and make sure nothing on it is a guide.
Page by page
| WordPress page | Decision | Why |
|---|---|---|
| Posts and pages | Index | The site |
| Homepage | Index | Obviously, and the first thing people get wrong when a plugin's "noindex archives" setting catches it |
| Hub pages (written, curated) | Index | The topic's main page; often the strongest page on the site |
| Category archives | Index if it's the hub; otherwise noindex | Two indexed pages for one topic compete |
| Tag archives | Noindex, or don't use tags | Thin, overlapping, one per keyword variant; the classic crawl leak |
| Date archives (year, month, day) | Noindex, or disable | Nobody searches your site by month |
| Author archives | Noindex on a one-author site; index on a multi-author site if the author page has a bio | A one-author archive duplicates the blog index; a real author page carries E-E-A-T |
| Attachment pages (one per uploaded image) | Redirect to the file, or noindex | A page with one image and no text; most SEO plugins have a one-click redirect |
| Internal search results (?s=) | Noindex and disallow in robots.txt | Infinite URLs; the one case for robots.txt |
| Pagination (/page/2/) | Index, self canonical | How Google reaches older posts; never canonical to page 1 |
| Thank-you, confirmation, unsubscribe pages | Noindex | Not for searchers |
| Privacy, disclosure, terms | Index | Harmless, expected, and reviewers look for them |
| Feeds (/feed/) | Leave alone | Google handles feeds; don't block them |
Noindex, robots.txt, canonical
| What it tells Google | Use it for | Don't use it for | |
|---|---|---|---|
| noindex (meta robots tag or X-Robots header) | "Read this page, but don't show it in results" | Removing thin pages from results while letting Google crawl through them | Pages you also block in robots.txt (Google can't see the tag) |
| robots.txt disallow | "Don't crawl this path" | Admin paths, internal search, infinite URL spaces | Removing a page from results; a blocked URL can still appear as a bare link |
| canonical | "This page is a copy of that one; credit that one" | Genuine duplicates (tracking parameters, print versions, www and non-www) | Pretending page 2 is page 1, or pointing different pages at one; Google ignores a canonical the content doesn't support |
The common mistake is combining the first two: a page noindexed and blocked in robots.txt stays in the index, because Google never gets to read the tag. Noindex needs the page crawlable. A second mistake is a canonical as a soft noindex; Google treats canonical as a hint and overrides it when the pages differ, so the "duplicate" stays indexed anyway.
Setting it
- In the SEO plugin's search appearance settings (Yoast, Rank Math and SEOPress all have them): per content type and per archive type, "show in search results" on or off. Turn off tags, date archives, author archive (one-author), and media/attachment (or choose "redirect attachment URLs to the file"). Leave posts, pages and the homepage on.
- Category archives: if the hub is a separate page, turn category archives off; if the hub is the category page, leave it on and write the introduction into the category description.
- Internal search: add
Disallow: /?s=andDisallow: /search/to robots.txt; the plugin usually noindexes the results template already. - Per page for thank-you and confirmation pages: the plugin's advanced tab on that page, "allow search engines to show this page" off.
- Don't touch pagination. If the plugin offers "noindex subpages of archives", leave it off, and make sure it isn't canonicalising to page 1.
Checking it
Search Console, Page indexing report, "Excluded by noindex tag": the list of URLs Google has seen the tag on. A month after the change it should hold the archives and nothing else.
Open it and search for /blog/ or your guide slugs; a guide on that list is a mistake, usually a per-page setting or a plugin rule that matched too widely. The not-indexed post covers the other exclusion statuses; this one is the only one you chose, so it's the only one that should contain only what you chose.
Noindexed by mistake
- The homepage, when it's set to show latest posts and the plugin treats it as an archive. Check it first after any archive change.
- The hub pages, when they're category pages and "noindex categories" was switched on for the thin ones.
- Every post, from the WordPress "discourage search engines" setting left on from development. It's one checkbox in Settings → Reading, and it's the most expensive checkbox in WordPress.
- Pagination, leaving posts only reachable from page 3 with no indexed path to them.
- A real author page on a multi-author site, where the bio and the list of articles are part of the site's expertise signal.
- Old posts someone decided were thin, instead of improving, merging or deleting them. Noindex isn't a content strategy; a page not worth indexing is usually not worth hosting.
Doing it with AffSERP
AffSERP publishes guides and informational posts as ordinary WordPress posts in the category you choose, so they fall on the "index" side of the table automatically and inherit whatever archive settings the site has.
Guides written by the X Best Review tool don't add tags, which keeps the tag-archive problem from growing with each batch, and the category each guide goes into is the one that maps to its hub.
The archive settings themselves are in the SEO plugin and are set once; the tools don't change them, which is the right division: the site decides what's indexable, the tools fill the indexable part.
FAQ
What pages should I noindex on a WordPress site?
The pages WordPress generates that have no content of their own: tag archives, date archives, attachment pages, internal search results, and author archives on a one-author site. Keep category archives indexed only if they act as hubs with their own text; otherwise noindex them too. Leave posts, pages, the real hubs and the homepage indexed, and leave pagination indexed but canonical to itself. The test for any page: would you want a searcher to land on it? If not, noindex.
What's the difference between noindex and robots.txt?
Noindex is a tag on the page that tells Google 'you may read this, but don't show it in results'; Google crawls the page, sees the tag, and drops it. Robots.txt blocks crawling: Google never reads the page, which means it never sees a noindex tag on it, and a page blocked in robots.txt can still appear in results as a bare URL if other pages link to it. To remove a page from results, use noindex and let Google crawl it. Use robots.txt only for things that should never be crawled at all, such as admin paths and internal search.
Should I noindex category pages?
Depends on what they are. A category archive that's just a list of excerpts with no text of its own is thin and competes with your hub page for the same topic; noindex it, or make the hub the category page. A category page with a written introduction, a curated list of the best guides and links down to each is a hub, and it should be indexed as the topic's main page. One or the other for each topic, never both indexed.
Should I noindex paginated pages?
No. Page 2 and onwards of a blog or category archive should stay indexed and carry a self-referencing canonical (not a canonical to page 1, which Google ignores and which hides the posts only reachable from page 2). Pagination is how Google finds older posts; noindexing it on a site with no other links to those posts can leave them orphaned. If the archives are noindexed as a whole, their pagination is too, which is fine as long as every post is linked from a hub or another post.
AffSERP publishes each guide as an ordinary post in its hub's category with no tags, so batches fill the indexable part of the site without growing the archives.
Start with free words