Content Audit for WordPress: How to Prune Posts That Drag Your Site Down

The Short Version

Start With Proof: CNET Deleted 3,000 Articles and Traffic Went Up 29%

CNET spent most of the 2010s publishing aggressively and accumulated a back catalog with thousands of pages that were outdated, thin, or competing with newer posts. It pruned around 3,000 of them — deleting or consolidating — and reported an organic traffic increase of roughly 29% afterward.

This is not an isolated outlier. A well-known marketplace removed 600,000 pages that received zero clicks over twelve months; clicks and impressions both rose about 30%. A smaller, documented case study reported a 32% traffic boost after removing low-value pages. And Semrush’s own analysis of pruning campaigns found content pruning alone driving an average 49% organic traffic increase.

Read those numbers again: deleting content raised traffic. Not adding content. Not rewriting content. Deleting.

That feels wrong until you understand how Google evaluates a site in 2026.

The helpful content system is no longer a separate filter — Google folded it into the core ranking algorithm in March 2024, which means site quality is now assessed at the domain level. A site where 30% of pages are thin doesn’t just rank worse on those 30% of pages. The algorithm assumes the rest probably shares the same problem, and suppresses the whole domain. When Google rolled out that change, it stated it was removing roughly 45% of what it considered unhelpful content from search results. Sites on the wrong side of that cut lost 30% to 90% of their organic traffic, and most never fully recovered.

Your thin pages are priced against your best pages. That’s the entire argument for a content audit, and it’s the reason you should treat this as a maintenance task, not an emergency response.

By mid-2025, independent SEOs began noticing Google quietly deindexing old thin pages, near-duplicate variants, and posts that simply paraphrased information already published elsewhere. Google doesn’t announce these sweeps. Your site can carry a hundred dead pages for years without you noticing — until one core update decides they’re part of the vote.

Why WordPress Sites Accumulate Junk Faster Than Most

WordPress makes publishing cheap, and cheap publishing leaves a paper trail:

• Tag and category archives — WordPress generates one automatically for every tag and category the moment you create it. A site with 300 tags has 300 (often contentless) archive pages
• Abandoned experiment posts — the “Wednesday Roundup” idea you published three times and dropped
• Recycled content — five posts on basically the same keyword, written months apart to chase rankings you never got
• Shortcode rot — content built with a plugin you deactivated two years ago now renders as raw text. WordPress doesn’t warn you; it silently breaks
• Version and pagination sprawl — /page/2 on archive pages, date-based archives, attachment pages your theme renders with one sentence
• The “it ranks for something” fallacy — a page pulling 40 visitors a month from a query nobody converts, indexed forever because nobody checked

None of this shows up in your dashboard. WordPress reports posts, not page quality. That’s the gap the audit fills.

Before you start, fix the two data-quality issues that will corrupt every number you look at: make sure your GA4 tracking isn’t counting your own visits, and check that your Search Console property is showing the domain correctly. A content audit built on polluted analytics is just another opinion. See how to exclude internal traffic in GA4 for WordPress if you suspect your data is dirty.

The Audit: Five Steps, One Spreadsheet

The output of every content audit should be a single spreadsheet — one row per URL — with enough columns to make the kill/keep decision without opening a single page. Here’s the workflow.

Step 1: Build the inventory

Export every indexable URL into one sheet. You need six columns minimum:

1. URL
2. Organic clicks — from Google Search Console, last 16 months (this window matches Google’s own reporting and flattens seasonal noise)
3. Impressions & average position — same source
4. Word count — export from your CMS or crawl
5. Last updated date — from your CMS
6. Referral/entry clicks and conversions — from GA4, last 12 months

Sources: Google Search Console’s Pages report gives you clicks, impressions, and position. GA4’s Pages and screenshields report gives you engagement and conversion data. A crawler (Screaming Frog free version handles a few hundred URLs; Ahrefs Site Audit and Semrush’s Content Audit pull in Search Console and GA4 data automatically) gives you word count, redirect status, and duplicate detection in one pass. If you don’t want a crawler, export posts with word count and dates via WP-CLI or a plugin and join them with the GSC export in a sheet.

The goal is one row per URL with all six values side by side. Stop when you can sort by clicks and immediately see the dead miles at the bottom.

Step 2: Score every page

You’re looking for six signals. A page qualifies as a pruning candidate when it hits one or more:

Word count is a tripwire, not a verdict. A 200-word page that crisply answers “how to change your WordPress theme” is fine. A 2,000-word page of filler that restates its title for two thousand words is thin. And since 2024, AI-generated filler with no original insight or data counts as thin too — length no longer excuses it.

If you want a shortcut for the mechanical checks (noindex flags, sub-300-word posts, missing meta descriptions, 18+ months stale, “discourage search engines” toggled on), a handful of WordPress plugins surface exactly these. The one that matters for most people already exists: same as your apostrophes. What you don’t need is another dashboard.

The critical step here is keyword cannibalization. Sort your search data by query instead of by page. When two or three of your posts rank for the same term, they’re splitting your clicks and confusing Google about which is the authority. The audit’s biggest wins usually come from collapsing these clusters, not from deleting zero-traffic junk.

Step 3: Decide — the only four real outcomes

Every page gets exactly one of these:

Keep. High-performing, current, no action. Do not touch what’s working.

Update. The page has a real position (or held one recently), gets traffic or converts, but the content is stale or thin. This is where the budget goes. Refresh facts, add depth, fix internal links, update the date.

Consolidate. The page overlaps with a stronger page — the classic cannibalization case. Merge the useful content into the survivor, then 301 redirect the loser to the winner. A 301 passes roughly 90–99% of link equity, so all the backlinks and authority the old page had now point at the page you kept. This single move fixes cannibalization, stops splitting equity, and strengthens your remaining content.

Delete. The page is irrelevant, has no backlinks, and removes value. Two options:

• 404 / delete — content you no longer host at all (expired promos, spam pages, defunct products) with no backlinks pointing at it
• 410 Gone — if a page has been archived intentionally and will never return, a 410 tells Google “gone for good” faster and more cleanly than a 404

What about pages that must stay live but shouldn’t rank — legal text, admin-ish utilities, “thank you” pages? Don’t delete what a human might legitimately need. Noindex it instead. That’s the one-click version of “keep but stop ranking,” and unlike a delete it keeps the URL working for anyone who has it bookmarked.

The discipline that makes or breaks the audit: decide ruthlessly, once, and write the decision next to the URL in the spreadsheet. Sites that shrink usually win. Sites that rationalize keeping everything get the same audit next year with a longer list.

Step 4: Execute in batches

Do not do this organically, one post every Tuesday. Batch it:

• Week 1–2: Audit complete, all decisions recorded
• Week 3–6: Execute improvements, merges, and deletions in batches — update first, then 301 redirects, then noindex, then deletes
• Month 2–3: Submit your sitemap again, resubmit the most important changed URLs in Search Console, and watch for re-crawls

After redirects and deletes, run a broken-link pass. Deleted pages referenced by surviving posts create 404s you now own. Fix internal links pointing at 301’d or deleted URLs, and run your crawler again to confirm zero broken internal links. If you hit redirect loops — the “too many redirects” error WordPress is famous for — you’ll be grateful you run these as batches rather than ad-hoc. See how to fix the too-many-redirects issue in WordPress before you start the redirect batch.

Two gotchas WordPress-specific: flush your rewrite rules / clear page cache after mass changes (or a cached old version of the file stays live for visitors), and remember that a 301 you set with a plugin years ago is still sitting there — audit your redirect map too, and delete any chain where A→B→C→D needs to just be A→D.

Step 5: Measure, then repeat

Pruning isn’t instant. Expect 60–90 days before ranking shifts on surviving pages show up in Search Console, and don’t interpret a bump in that window as proof — core updates may land in the same period. The honest test is comparative: track the surviving pages’ traffic, positions, and crawl frequency against the quarter before the audit.

Then put it on a schedule. Quarterly for the mechanical checks (thin content, noindex, stale), annually for the full strategic pass. An audit done once is a tune-up; an audit you repeat is a quality system.

What to Prune When You Can’t Decide

If a page feels borderline, run these three filters in order:

1. Does it have backlinks? Any external links pointing to it are equity you’ll either keep (update/consolidate) or lose (delete). Check Site Explorer or comparable backlink data before choosing delete.
2. Does it convert, or has it ever? A page with traffic but no conversions and a page with conversions but no traffic are both problems, but opposite ones.
3. If it were unpublished tomorrow, would anyone notice? Honestly, no. That’s your answer.

What a Content Audit Doesn’t Do

Be honest about the limits, or you’ll oversell the effort to yourself:

• It won’t fix a manual action. If Search Console shows a penalty, pruning won’t remove it. Fix the cause first, then prune.
• It won’t make a weak page strong. Culling releases the site-wide drag. The growth after a good audit comes from the strong pages you keep and from new content published afterward — both pull from the same rebuilt quality score.
• It won’t survive a growth plan that keeps producing thin pages. If your publishing cadence is the problem, the audit is a band-aid. Throttle the publishing.

WordPress-Specific Junk Most Audits Miss

Add these to your checklist — generic content audits don’t look for them, and WordPress generates them by default:

• Default archives — tag, category, author, and date archives. Disable the ones you never fill, noindex the ones you keep
• Attachment pages — themes render these with one sentence and an image. Point them at the parent post or noindex them
• Pagination dupes — /page/2 and beyond on thin archive pages. Consolidate or noindex
• Old theme cleanup — posts with shortcodes or blocks from themes/plugins you’ve since removed
• Test/staging leaks — posts published from a staging environment that never should have gone live

Keep your robots.txt honest while you’re at it. Blocking crawler access to pages you also list in your sitemap is a contradiction that sends mixed signals — which is why stopping search engines from crawling WordPress is worth a deliberate decision, not a default.

FAQ

Does deleting blog posts hurt SEO?

No — if you do it with a 301 redirect for pages that have backlinks or traffic. Deleting a page with no backlinks and no traffic removes a drag on your domain. Deleting a page with links and redirecting it to a related page moves its equity to a stronger URL. The mistake is deleting useful pages outright (404) or deleting linked pages without a redirect.

How much content should I prune?

There’s no percentage target. Clean evidence outranks volume — a small set of strong pages beats a large set of thin ones. Start with zero-click pages that have no backlinks, then cannibalized clusters, then genuinely outdated posts. If you’ve never audited, expect to retire or merge more pages than you keep.

What’s the difference between noindex and deleting a page?

Noindex keeps the page live and crawlable but tells Google not to show it in search results. Deleting removes it entirely. Use noindex for pages humans may legitimately reach (legal pages, account utilities, thank-you pages). Delete (or 404/410) pages that should not exist or be reachable.

How long does it take to see results after pruning?

Generally 60–90 days for ranking movement on surviving pages, per case study data — longer if the recrawl queue is slow on a large site. Traffic bumps can also be delayed if you’re pruning during a core update, so judge the outcome over a full quarter against the previous baseline.

What’s the best tool for a WordPress content audit?

There’s no single best tool; the workflow beats the tool. Google Search Console (free) gives clicks, impressions, and position. GA4 gives engagement and conversions. A crawler like Screaming Frog (free up to a few hundred URLs), Ahrefs Site Audit, or Semrush Content Audit fills in word counts and technical issues. WordPress plugins surface the mechanical checks (thin, stale, noindexed) inside your dashboard. Combine at least GSC + a crawl — one source is never enough to trust a delete decision.

Does AI search care about pruning?

Indirectly, yes. If you’ve published a llms.txt for WordPress or are optimizing for AI visibility, a curated map of your site means little if it lists a hundred pages you should have retired. The same site-quality logic applies: agents and AI search tools favor sites they can parse quickly, and a small set of strong, clean pages is easier to serve than a sprawling one with dead ends.

Should You Bother?

Here’s the honest framing: if you’ve been publishing WordPress content for more than a year and have never run an audit, the odds are good that more than a quarter of your indexed pages are costing you search visibility — not because Google penalizes them individually, but because site-wide quality assessment means your thin pages vote against your strong ones.

An audit is not a content marketing trend. It’s a maintenance task with a predictable payoff, backed by the oldest logic in SEO: the algorithm assesses the domain, and the domain is only as strong as its worst indexed pages. Merge the overlaps, 301 the cannibals, fix what has value, noindex what must stay, and delete what earns nothing — then wait a quarter and watch what the survivors do.

And build the habit. The sites that win after core updates aren’t the ones that knew what to do in panic week; they’re the ones that ran the audit in March when nothing was wrong.