Rankings dropped ~40% after the June 2026 core update — what's the actual recovery playbook?
Site has been stable for two years. The June 9 core update hit and organic is down ~40% sitewide, worst on our money pages. GSC shows positions slipping from 3–5 to 12–18 on a big chunk of head terms.
Nothing technical changed — no migration, no CMS update, robots.txt untouched. Before I start hacking at content, what's the disciplined order of operations people actually use to diagnose and recover from a core update? Trying not to panic-edit.
Resist the urge to make sweeping changes in week one. Core updates are reassessments of relevance and quality, not penalties — so there's nothing to 'fix' in the technical sense. Here's the order I run every time:
1) Segment the loss. In GSC, compare the 28 days before vs after by page and query. Sitewide-even usually means a quality/trust reassessment; section-specific points at thin or outdated clusters.
2) Map intent shifts. Pull the top 20 fallen queries and look at the live SERP. Updates often change what intent Google rewards. If the SERP is now all forums and yours is a thin affiliate page, no on-page tweak saves it.
3) Audit your worst losers against the helpful-content self-assessment honestly. Cut or consolidate the genuinely weak stuff — pruning lifts the whole domain's average.
4) Then improve, don't churn. Add real depth, original data, author credibility to the survivors. Ship in batches and wait. Recovery typically lands on the NEXT core update, not in days.
Following exactly this. We lost 35% in March, pruned ~400 thin pages, recovered +20% on the June update. Pruning was the single biggest lever.
One thing people miss: check if it's a classic core update or a spam/reviews component rolling at the same time. Google bundled several in June — different problems, different fixes.
Do NOT disavow links in a panic. Core updates are almost never about links and a bad disavow file can do real damage if you're wrong.
Look at branded vs non-branded too. If branded held and non-branded tanked, it's a relevance reassessment, not a trust/EEAT hit.
We tested 'refresh dates only' vs 'genuine content depth' on two cohorts. Date-only did nothing. Depth moved it. Stop gaming freshness.
New here — how long is 'wait'? My boss wants results this week and I'm trying to set expectations.
@Mia realistically until the next confirmed core update, historically 4–8 weeks out. Set that expectation now or you'll get pushed into churn that hurts.
Underrated move: interview a real SME and add their quotes + bio. The EEAT 'experience' signal is doing heavy lifting post-2025.
Check Discover and Top Stories separately — those have their own systems and can mask or exaggerate the core-update delta in aggregate reports.
Anyone else seeing forums and Reddit eat the spots? Half my lost queries are now UGC results. Hard to beat with on-page edits alone.
@Mason yes — that's the intent shift in #2. If Google decided the query wants peer discussion, the answer is a community/Q&A format, not another guide.
We added a genuine 'who we are / why trust us' section + author pages with real credentials. Recovered the YMYL pages specifically.
Save your GSC data weekly. The 16-month window will roll off your pre-update baseline and you'll wish you'd kept it.
Counterpoint to pruning: don't delete blindly — consolidate with redirects. I nuked pages that had useful inbound links and lost that equity.
Patience is the whole game. I've watched three clients recover and every single one was on the following update, never in between.
Bookmarking this. The 'segment first, edit last' framing should be pinned somewhere permanent.