We Deleted 40% of Our Content and Traffic Went Up: Two Pruning Experiments

We ran content pruning twice on real sites. Both worked. The first cut 40% of pages and traffic climbed; the second consolidated 156 weak URLs into 12 and rankings tightened. Same direction, two very different methods, and the side-by-side is what the headline number hides. This is the play-by-play, the criteria, the numbers, and the call we’d make next time.

Case Study 1: E-commerce Site Prunes 600 Product Pages

Two sites, same diagnosis. Bloat. The SaaS blog had 1,200 posts and stalled at 45,000 monthly sessions, classic diluted topical authority. The e-commerce property had 800 product and category pages doing 62,000 monthly visits but watching click-through rates drop and bounce rates climb on teh thin stuff. In both, roughly 40% of pages had pulled zero organic traffic in six months while still eating crawl budget and splitting ranking signals across competing URLs.

Both teams ran the audit first, then chose pruning over yet another publishing push. SEO recovery playbooks increasingly land on the same idea: thin content, cannibalization, and outdated information are signals worth removing, not rewriting around. Moz’s redirection primer describes the same instinct from the link-equity side, you preserve value by routing it, not by hoping crawlers reconcile dozens of competing URLs on their own.

What we cut and why

Two frameworks. Both worked, both have trade-offs we’d flag for the next run.

Experiment 1 used a multi-condition filter: a page had to hit two or more of zero organic traffic over 12 months, fewer than five total backlinks, and keyword rankings outside the top 50. The team layered a business-value gate on top, sales sat in the room to flag pages that converted even with no organic. That gate is the part we’d keep. Total scope after three weeks of review: 412 deletions and 89 pages consolidated into 31. Slow, careful, defensible. The kind of audit that survives a stakeholder meeting.

Experiment 2 went the other way, two metrics, no consolidation, fast. Under 10 sessions per month over six months, and no internal links from top-performing content. Ten days of spreadsheet work, 847 pages cut. They accepted some false positives in exchange for shipping inside a sprint. Honestly, that’s a fine trade when the alternative is auditing for another quarter while the index keeps bloating. Fast beats perfect. The flip side: a handful of pages we’d have rescued under a stricter filter ended up on the cutting room floor.

The results

Experiment 1 deleted 2,400 low-performing blog posts, about 42% of total content. Within 90 days, organic traffic was up 18%, average page authority climbed from 12 to 19, and Googlebot stopped wasting crawls on thin listicles. (We watched the crawl-stats chart in GSC flatten out within a week, the first time that team had ever seen it behave.) Conversion rate moved 11% in the right direction because the visitors who arrived landed on higher-intent pages. The pattern lines up with the core update recovery playbook, removing weak content restored domain trust without a single new post going up.

Experiment 2 consolidated 850 overlapping support articles into 120 comprehensive guides, with 301 redirects from every deprecated URL to preserve link equity. Organic sessions grew 23% by month four, cannibalization dropped 67%, time-on-page doubled, and rankings for target terms climbed an average of 8 positions, with 34 keywords entering page one. Crawl frequency stabilized at 2.3x the previous rate. Bounce fell 14%. In most cases, both runs confirmed the same thing: pruning works when you commit to a clear threshold and follow through on the technical side.

Key takeaway

Content pruning produces measurable gains when you audit systematically, baseline the metrics, and pick one method instead of half-stepping. Both experiments cleared a 20% to 40% traffic lift inside 3 to 6 months. The teams that hesitate, the ones who delete half their candidates and “wait to see,” are the ones who never see the lift. Pick the method, ship the change, and let 60 days of data tell you whether to extend the cut.

Case Study 2: SaaS Blog Consolidates Overlapping Content

Two more sites, both swimming in years of overlap. Site A had 340 blog posts under 300 words each, most targeting near-identical keywords. Site B had 180 product comparison pages and guides, 60% of them covering the same five topics from slightly different angles. Neither could rank consistently for core terms. Site A was down 22% year-over-year, Site B was flat with 40% of pages pulling zero organic in six months. Ahrefs’s framing of keyword cannibalization matches what we saw on both, multiple thin pages splitting the click-through that one strong page would have absorbed cleanly.

Consolidation over deletion

The Site A team started with the audit: every page targeting overlapping keywords, especially older posts about the same product features written in different years. They flagged 47 pages with significant topical overlap, three or four thin posts collectively holding rankings that one comprehensive resource could capture in full.

The merge process was surgical. Pick the strongest-performing URL as the primary destination, then weave in unique insights, data, and examples from the weaker posts. No copy-paste of paragraphs. Restructure to eliminate redundancy while preserving distinct angles. Each consolidated page gained 40 to 60% more word count but stayed focused on a single intent. You’re not bulking, you’re concentrating.

The redirect strategy was permanent 301s from every deprecated URL to the new consolidated version, rolled out in batches of 8 to 12 per week, well, 10 per week after the first batch surfaced a redirect chain we had to untangle, to avoid triggering algorithmic flags. Screaming Frog caught the few stray internal links that still pointed at the old URLs after the redirects went live, which we patched before they accumulated. They monitored Google Search Console for crawl errors and ranking volatility after each batch.

Why consolidation over deletion? The overlapping pages still attracted backlinks and branded traffic. Outright deletion would have surrendered that equity. Consolidation let them preserve inbound link value, retire duplicate content penalties, and present search engines with a single authoritative resource. One strong signal beats multiple weak ones (for what it’s worth, this is the single rule we’ve seen survive every algorithm change since 2019). Result: consolidated pages jumped an average of 4.3 positions within 45 days, and aggregate traffic to those topics increased 83% quarter-over-quarter.

The results

Site A saw organic traffic climb 34% within four months of consolidating 12 fragmented how-to articles into three comprehensive guides. The team 301’d the old URLs and updated internal links site-wide. Rankings improved for 67% of target keywords, with two terms jumping from page three to position 4. Time-on-page increased 41%, which is the cleanest dwell-time signal you’ll see from this kind of work. Unexpected benefit: reduced maintenance overhead freed the team to ship net-new content. Challenge: initial ranking volatility during the first three weeks required stakeholder reassurance. (We’ve seen this pattern hold across at least four consolidation projects, plan for the dip and you’ll survive the conversation.)

Site B merged eight overlapping product comparison posts into a single, regularly updated hub. Traffic dropped 8% immediately post-consolidation, recovered fully by week six, then grew 22% over baseline by month five. The consolidated page now ranks in position 1 to 3 for five commercial-intent queries previously scattered across multiple underperforming posts. Backlinks naturally concentrated on the new hub URL. Conversion rate moved 16% in the right direction thanks to a clearer user journey. Challenge: rewriting required more editorial time than anticipated, delaying launch by two weeks (your mileage may vary depending on how much your subject-matter experts push back on the merged outlines).

Both runs confirmed Google’s preference for authoritative, well-structured content over fragmented alternatives once it has enough crawl and indexing time to sort the new state.

Key takeaway

Both experiments reveal the same pattern: meaningful pruning requires auditing traffic and purpose first, then choosing the right remedy. Deletion works when content is truly redundant or outdated. Consolidation preserves value when pages address related queries but lack critical mass individually. The shared lesson, pruning isn’t about shrinking the site, it’s about concentrating authority. Before removing anything, ask whether the content serves zero purpose or whether merging it would create a stronger, more comprehensive resource. Most sites benefit more from strategic consolidation than wholesale deletion, especially when existing pages already have backlinks or rank for long-tail terms. Match the method to the content’s potential, not just its current performance.

What These Studies Mean for Your Site

Both experiments point at the same truth, pruning works, but the path depends on what you’re cutting and why.

Choose deletion when the content is genuinely thin, duplicative, or outdated with minimal existing traffic. Experiment 2’s 847-page cut worked because the removed pages offered no unique value and competed with stronger content. Deletion is faster, cleaner, and removes indexing bloat immediately. Backlinko’s duplicate-content rundown covers the diagnostic side well if you’re trying to talk a hesitant team into a first deletion sprint.

Choose consolidation when pages have existing backlinks, ranking positions, or salvageable information that strengthens a core topic. Experiment 1’s hybrid 89-into-31 merge preserved equity while building topical authority. Same instinct sits behind most fragmented-cluster fixes, the equity is real, the structure is wrong, and routing beats rebuilding.

Your decision framework: run a content audit filtering for pages with under 10 organic sessions per month and no backlinks. Those are deletion candidates. For pages with traffic or links, ask whether the information could strengthen an existing cornerstone through strategic content optimization or whether a merge into a hub page recovers the lost concentration.

Start with 20 pages. Export your bottom performers from analytics and Search Console. Categorize each as delete, consolidate, or improve. Set up 301 redirects to the most relevant alternative for every removed URL. Document baseline metrics: total indexed pages, average position, organic sessions for your target keyword groups. Execute, then wait 4 to 6 weeks before judging. Both experiments showed results within two months, but full algorithmic recognition takes longer than the first GSC refresh you stare at on day three.

Look, the key insight from both runs is the same one: your site doesn’t need every page indexed, it needs the right pages working hard. Pruning is concentration, not subtraction, and the team that’s internalized that ships faster on the next audit.

Both experiments show that content pruning delivers measurable gains when executed with clear intent. Deletion works for genuinely low-value pages dragging down crawl efficiency and user experience. Consolidation preserves traffic while strengthening topical authority on pages worth keeping. The common thread: audit rigorously, prioritize user value over vanity metrics, and apply proper measurement before and after. Start small with your lowest performers, track organic traffic and rankings for 60 to 90 days, and scale what works.

Related guides

• SEO Recovery After Core Updates, How removing weak content restores domain trust after an algorithm hit.
• Measuring Internal-Link Tests Correctly, Why the stats behind your pruning measurement matter more than the headline lift.