A Google core update hit your site: what it means and what to do
You saw a drop and a core update rolled out. Here is what a core update is, how to confirm the cause, and what to do next. Built for founders who ship the site themselves.
By Théophile Louvart, founder of Porteur · Updated 14 September 2026 · Markdown
What a core update is and where to check the date
A core update is a broad change to Google’s ranking systems. It happens several times a year, as of 2026.
Google announces core updates on the Search Status Dashboard. Each has a start and an end. A rollout takes about two weeks.
Find the announcement
Open status.search.google.com and switch to the Ranking tab. Look for “Core update” with a start date and an “Ended” date.
Match the window
Compare the rollout dates with your Search Console Performance report’s date range. Set a filter to the same window.
Note other updates
Check if a spam update also ran. If yes, treat that as a separate line of inquiry later in this guide.
If you launched a big change on your site at the same time, note that too. Do not assume the update alone caused the change.
Confirm it was the update, not your changes or a bug
You need a date match and a pattern match. Core updates move groups of queries and pages at once. Random single-page dips are less likely to be core-wide.
- Site-wide pattern: traffic from many unrelated queries falls within the rollout window.
- Query shuffle: rankings for “project management templates” drop while “kanban board app” rises, across multiple pages.
- Stable tech: no deploys, no outages, no robots.txt or noindex changes in the same week.
- Search Console clean: no Manual actions and no Security issues in the same period.
Check Manual actions
In Search Console, open Manual actions. If you see an item, this is not a core update issue. Fix and file a reconsideration request.
Check Security issues
Open Security issues. If you see hacked or malware labels, treat them first. These are separate from core updates.
Compare pages vs queries
In the Performance report, switch between Pages and Queries. If most top pages and most top queries fell together, it fits a core update.
Exclude changes you made
Review your deploy log. If you changed /pricing or noindexed /blog/* that week, isolate the effect with annotations and page-level graphs.
A clean account with a clear date and site-wide pattern points to the core update. Move on to content assessment next.
Core update vs spam update vs manual action
Core updates reassess content across the web. Spam updates enforce spam policies. Manual actions are human reviews that find a violation on your site.
| Thing | How it’s announced | What it targets | Where you see it | What to do |
|---|---|---|---|---|
| Core update | Search Status Dashboard, start and end dates | Quality, originality, usefulness and E-E-A-T signals across the index | Traffic and rankings change, no alert in your account | Reassess content, improve pages, wait for reprocessing |
| Spam update | Search Status Dashboard, start and end dates | Practices in spam policies such as cloaking, doorway pages, scraped or scaled content abuse, link spam | Traffic can drop. No account alert unless also a manual action | Stop the behaviour, remove or noindex spammed pages, clean up links if you bought or sold them |
| Manual action | Only in Search Console’s Manual actions report | A specific violation on specific pages or site-wide | A notice with the type, examples and scope | Fix fully, document, submit a reconsideration request and wait days to weeks |
The helpful content system was folded into the core systems in March 2024. Its signals are now part of core updates as of 2026.
What Google says to do after a drop
Google’s guidance: there is nothing to fix for a page that dropped in a core update. The systems reassessed content against other content on the web.
- Review quality, originality and usefulness. Would a user bookmark this page or send it to a colleague?
- Check expertise, experience, authoritativeness and trust. Show first-hand use, real names and verifiable claims.
- Remove or improve weak pages that pull the topic down.
- Strengthen internal links to your best answers on the topic.
Recovery, when it comes, often waits for a later update. Set expectations now: weeks to months, not days. Keep shipping better pages in the meantime.
Work through the content questions, page by page
Start with the pages that lost the most clicks. Use your Performance report to rank by lost clicks in the 28 days after the rollout against the 28 days before.
Match search intent again
For /guides/getting-started that ranks for “how to set up time tracking”, add a clear step-by-step with screenshots and a video. Remove fluff.
Add first-hand experience
If you compare tools on /blog/best-kanban-tools, disclose your criteria, show your own test notes, and include failures or trade-offs.
Tighten the scope
Split catch-all pages. If /blog/google-forms-guide tries to cover forms, sheets and sites, break it into separate guides and interlink them.
Improve originality
Replace generic lists with your own data or examples. For “invoice template” add a downloadable yourproduct.com/invoice.xlsx and a worked example.
Make expertise visible
Add bylines with roles, link to author pages, and cite sources. Add a last reviewed date when you genuinely reviewed and updated the page.
Raise usefulness
Embed a calculator, a copyable command or a live demo. For /pricing, add a toggle for annual vs monthly and a usage slider tied to your product.
A fixed page looks like this: fewer headings, clearer answers near the top, concrete examples, and links to deeper pages you control, not to thin hubs.
- Remove spun summaries or AI rewrites that add nothing new.
- Cut thin affiliate sections that only restate a vendor’s features.
- Noindex pages that cannot be improved this quarter.
- Merge duplicates, keep the strongest URL, 301 the rest.
Site-level patterns to check before rewriting everything
Core updates often reflect site-wide patterns. Fix those first, then go page by page. This saves time on a small team.
- Topic scope: if you publish about everything, narrow to what you know. Build clusters, like /guides/forms/* with five deep pages, not fifty shallow.
- Internal links: make sure your best pages are two clicks from the home page. Link from /guides/google-forms to /guides/google-sheets if users need both.
- Authorship: add author pages, credentials, and a short “Why trust us” on evergreen guides.
- Out-of-date content: run a content audit. Update high-potential pages with decayed traffic. Remove or redirect the dead ones.
- UX basics: keep LCP under 2.5 seconds, CLS under 0.1, and INP under 200 ms. Bad UX weakens usefulness even if content is good.
Do not confuse this with a technical SEO fire drill. Fix crawl barriers if you have them, but core updates are about content and overall value first.
Timeline of recovery and what to watch
Expect a bumpy graph. You may see partial gains before the end of the rollout, then a plateau. Larger recoveries tend to align with later core updates.
Week 0 to 2
Rollout window. Do not thrash. Gather deltas, mark the dates, shortlist affected topics and pages.
Week 2 to 6
Ship concrete improvements on five to ten pages. Add experience, tighten scope, and improve internal links from related pages.
Month 2 to 4
Expand the programme to 20 to 30 pages if you have them. Publish new, original pages in the same clusters.
Next core update
Expect the bigger reassessment here. Keep tracking queries by topic cluster, not just site totals.
- Track clicks and average position per topic, not just all pages.
- Watch which queries return. Sometimes new queries replace the old ones.
- Compare against rivals on your main terms. See who rose, study their pages’ structure and signals.
What not to do
- Do not mass rewrite every page in a week. You will lower quality and lose your voice.
- Do not stuff keywords to claw back positions. It weakens usefulness and can trip spam systems for keyword stuffing.
- Do not buy links or join link farms to offset the drop. Link spam is in the spam policies and can lead to a manual action.
- Do not file a disavow file because a tool flagged “toxic backlinks”. Google says most sites do not need it and they already ignore most bad links.
- Do not switch domains to escape a drop. You will reset signals and risk a deeper loss.
Links, sandbox talk and negative SEO
‘Toxic backlinks’ is a term from link tools, not from Google. Google’s representatives say those toxicity scores are not Google’s. Disavowing on their basis is unnecessary for most sites as of 2026.
Negative SEO through links is rare to have an effect. Google says its systems ignore such links in nearly all cases. Spend time on content and UX instead.
The ‘sandbox’ is a long-standing observation that new sites take months to rank on competitive queries. Google has not confirmed a sandbox mechanism. The delay fits the time its systems need to gather signals about a new site.
A practical recovery plan for a small site
Triage the top losses
In Search Console, export pages with the largest 28-day click losses. Keep the top 20. Group by topic, not just by URL.
Decide the action per page
Keep and improve, merge into a stronger URL, or noindex and park. For duplicates, keep /guides/google-forms, 301 /blog/google-forms-guide to it.
Raise proof and usefulness
Add your own data, screenshots, code, or templates. Replace stock images. For /guides/forms-vs-sheets, add a decision table and a downloadable checklist.
Tighten internal links
From /guides/getting-started link to /guides/forms-templates and /guides/form-responses. Use descriptive anchors, not “click here”.
Publish one new standout page
Fill a gap you can win. If rivals do “google chat” basics, publish “google chat for customer support teams” with real workflows.
Review again after 28 days
Measure clicks and positions. If a page shows no movement and is weak, consider merging or noindexing it until you can improve it.
Keep scope tight. Five great pages on one problem beat fifty summaries across unrelated tools like Sheets, Forms and Sites all at once.
Questions
A core update is a broad change to Google’s ranking systems. It happens several times a year. It reassesses content quality, originality, usefulness and signals like E-E-A-T across the web. It is announced on the Search Status Dashboard with a start and end date.
Check the Search Status Dashboard. It lists recent core updates with start and end dates. A rollout takes about two weeks. Match those dates to your traffic changes to confirm impact.
There is no fixed timeline. You can see small gains within weeks if you improve key pages. Larger recoveries often align with later core updates, when systems reassess more broadly. Plan for weeks to months.
No. Do not mass rewrite. Focus on pages that lost the most and matter to your product. Improve usefulness and first-hand experience, merge duplicates, and noindex what you cannot fix now.
Usually no. Google says most sites do not need it and they already ignore most bad links. Only consider it if you have a manual action for link spam or a history of paid links you cannot remove. A disavow file cannot be undone quickly.
It could. Check Search Console’s Manual actions and Security issues. Spam updates also roll out and target policy violations like link spam or scaled content abuse. If you have a manual action, fix it and file a reconsideration request.
Sources
Check my site, free
Drop your site URL to see three free findings on where you lost and which rivals rose on your terms, in about thirty seconds, then decide if Porteur is worth a closer look.
- Free check, no card
- Read-only, your own accounts
- Readable by your agent
Read next
- GuideGoogle spam updates: the policies they enforce and how to stay clear
- GuideManual actions and the reconsideration request
- GuideThe disavow tool: what it is for, and why you almost never need it
- GuideToxic backlinks: a tool’s word, and what Google actually says
- GuideHow to find declining pages in Search Console before the traffic is gone
- GuideA content audit for a small site: one table, one decision a row
- GlossaryRanking
- GuideHow to submit a website to Google in 2026 (and why you mostly do not have to)