Manual action for pure spam — diagnose and recover

Search Console's Manual Actions report shows a 'Pure Spam' or 'Thin content' entry with either a site-wide or partial-match scope. Partial matches name affected URL patterns; site-wide matches do not. Within a day or two of the action landing, the affected templates fall out of the top 100 entirely — this is suppression, not a ranking adjustment, so positions go to null rather than dropping a few places. Branded queries usually survive; everything non-branded on the flagged templates evaporates. Unlike an algorithmic hit, you get an explicit notification and a reconsideration channel, which means recovery is gated by a human reviewer rather than the next core update.

Programmatic pages with no information a human would value

Pure Spam actions land on templates where the generated pages exist only to capture a keyword permutation — entity-swap pages, auto-generated location or category pages with no unique data, or scraped-then-spun content. The reviewer's test is whether the page would exist if search engines did not. If the honest answer is no, the action stands until those pages are gone or genuinely rebuilt.

Scaled content abuse: thousands of near-identical pages shipped fast

The March 2024 scaled-content-abuse policy made volume itself a signal. A site that adds 20,000 templated pages in a quarter, each sharing >70% boilerplate, reads as an abuse pattern even if any single page looks borderline acceptable. Reviewers see the aggregate, not the individual URL, so a fix that touches only a handful of sample pages will not clear the action.

Sneaky redirects, cloaking, or doorway sets feeding a money page

Doorway pages — large sets of similar pages that all funnel the user to the same destination — are an explicit Pure Spam trigger. If your programmatic pages exist mainly to rank and then push the visitor to a single conversion page, the reviewer will classify the whole set as doorways regardless of how polished each page looks.

Expired-domain or subfolder abuse inherited from a prior owner

If the domain or a subfolder was previously used for spam, the manual action can attach to history you did not create. This is common after acquisitions. The fix is the same — remove or rebuild the offending content — but the reconsideration request must explicitly document the ownership change and what was removed.

1. 1

   Open Search Console → Security & Manual Actions → Manual Actions. Record the exact label ('Pure Spam' vs 'Thin content') and whether the scope is site-wide or partial — partial actions list the affected URL patterns you must focus on.

2. 2

   Map the named patterns (or, for site-wide, your largest templates) to path prefixes and parameter shapes so you know precisely which page sets the reviewer is judging.

3. 3

   Run pseolint against the affected templates and read the spam-cluster findings first — doorway-pattern, near-duplicate, thin-content, and boilerplate-ratio are the rules that correspond to Pure Spam reasoning.

4. 4

   Sample 25 random URLs per affected template and apply the human test to each: would this page exist if search did not? Tally the pass rate — below ~70% means the template needs rebuilding, not editing.

5. 5

   Decide per template: rebuild with genuine per-page data, consolidate many thin pages into fewer substantive ones, or remove and 410. Do not noindex-and-leave; reviewers want the spam gone, not merely hidden.

6. 6

   Ship the fixes site-wide before filing — a reconsideration filed while spam still exists on unflagged templates is the most common rejection reason.

7. 7

   File the reconsideration request with a specific, honest writeup: what was wrong, the scope of pages changed, links to before/after examples, and how you will prevent recurrence. Vague requests are rejected and reset the queue.

pseolint findings most strongly correlated with this pattern.

A travel-deals site received a site-wide Pure Spam action covering ~40,000 auto-generated '{city} cheap flights' pages that each wrapped an affiliate widget in spun boilerplate. The team deleted 38,000 of them with 410s, rebuilt 2,000 top-demand routes with real fare-history data, median prices, and editorial route notes, and removed two doorway funnel pages. Their first reconsideration was rejected for residual thin pages on a forgotten subdomain; the second, filed 18 days later after cleaning the subdomain, was approved. Non-branded organic recovered to 58% of pre-action levels over the following 90 days as the rebuilt pages re-earned rankings on merit.

Recovery has two clocks. The first is the reconsideration cycle: one to three weeks per submission, with rejections resetting the queue — so realistic clearance is two to six weeks if your first filing is thorough, longer if you iterate. The second is the ranking clock: once the action lifts, the surviving pages must re-earn position on merit, which lands over the following 30 to 90 days as Google recrawls and rescores. Do not expect a snap-back to pre-action traffic; the spam pages that produced most of that traffic are gone by design. Track the Manual Actions report for the 'no issues detected' state, then watch the per-template Performance report for the recrawl-driven climb.