Fixing a Ranking Drop After a Google Core Update
A core update is a broad change to how Google weighs content across the whole index. When one rolls out, some pages rise and others fall, often without anything having changed on the sites involved.
If your traffic dropped during a rollout, the hard part is working out what the loss actually is: which pages, which queries, how far, and what now ranks instead. This guide takes you from a worrying traffic chart to a measured list of losing pages, an honest comparison with the winners, and a realistic improvement plan you can track through the next update.
What a core update actually changes
Google runs many ranking systems at once. A core update adjusts how those systems assess and weigh content in general, rather than targeting one practice such as bought links. Google usually announces core updates on its Search Status Dashboard with a start date, and posts again when the rollout has finished. Rollouts have commonly taken around two weeks, though some have run longer.
Three things follow from the way core updates work, and they shape everything in this guide:
- There is no single error to correct. Google's own guidance says a page that drops has not necessarily done anything wrong. Other pages may simply now be judged more useful for the same searches.
- The assessment is relative. Your page is weighed against everything else that could answer the query. A page that has not changed in three years can lose ground purely because better answers now exist.
- Recognition takes time. Improvements you make are picked up as pages are recrawled and reassessed. Google has said some change can appear between updates, but broad quality improvements are often most visible when a later core update runs.
Since March 2024, the signals that the separate helpful content system used to provide are part of the core ranking systems. That means a core update can now reflect a sitewide judgement about how much of your content seems written for people rather than for search engines. If your whole site slid rather than a handful of pages, read the helpful content guide alongside this one.
Reading the drop in Search Console
The Performance report in Search Console is the most direct evidence you have, because it records how Google actually showed your pages. Set it up like this:
- Open Performance, then Search results, and tick Average position as well as clicks and impressions.
- Choose Date, then Compare, and set a custom range: a period after the rollout ended against a period of the same length before it started. Use whole weeks so weekday and weekend patterns balance out.
- Open the Pages tab and sort by the difference in clicks. Export the table.
- Repeat on the Queries tab, and again with the Search appearance tab if you had rich results.
In your spreadsheet, work out the change in average position for each page, not just the change in clicks. That lets you sort pages into patterns, each pointing in a different direction:
| Pattern | What it usually means | Next move |
|---|---|---|
| Small slips, for example position 2 to position 4 | Minor reordering among similar results. Often not worth a rewrite. | Watch for a few weeks; improve only if the page is genuinely weak. |
| Large falls, for example top three to page two or lower | Google now judges other pages far more useful for that query. | Full comparison with the new winners, then substantial improvement. |
| Most pages down by a similar amount | A sitewide assessment has shifted, not a page-level one. | Review the site as a whole, including thin and unhelpful sections. |
| One section down, the rest steady | That type of content, such as reviews, guides or category pages, fell out of favour. | Focus on that template and its content model. |
| Positions steady, clicks down | Not a ranking loss. Demand, seasonality or result layout changed. | Check search features such as AI Overviews above your listing. |
Keep the export. You will want the same list months later to see whether the pages you improved recovered faster than the ones you left alone.
Comparing yourself with the pages that gained
The most useful evidence after a core update is not in your own analytics. It is on the results page. For each query where you lost significantly, open a private browser window, search, and note the pages now in the top five. Then ask specific questions and record the answers in columns:
- Who wrote it? Is there a named person with relevant experience, or a recognised organisation behind it?
- What does it have that you do not? Original photos, test results, prices, worked examples, a tool, local knowledge, first-hand accounts?
- What kind of page is it? If you wrote a long guide and the winners are short answers, product pages or forum threads, the intent Google serves has changed.
- How current is it? Does it reflect how things work now, or is it the same age as yours?
- How easy is it to use? Does the answer appear quickly, or is it buried under introductions and advertising?
Look for what the winners share, not what one of them happens to do. If four of the top five are run by specialists and include original testing, that is a signal. If one has a long word count and the others are short, word count is not the signal. Google has said there is no minimum or ideal length.
When the winners are forums and video
For many queries Google now shows discussion forums, video results and AI-generated summaries more prominently than before. If your page was replaced by a Reddit thread or a video, the lesson is usually that searchers wanted lived experience or demonstration, not a generic article. Adding real experience to your page, or answering a narrower question better, is more promising than trying to outrank the forum on its own terms.
The fix, step by step
Step 1
Let the rollout finish and confirm the timing
Check the Google Search Status Dashboard for the start and completion dates of the update, and do not rewrite anything while it is still rolling out, because positions often swing during that period and settle afterwards. Open the Algorithm Impact monitor to line up your daily ranking movement with the update timeline, and note whether your decline began inside the rollout window or before it started.
Step 2
Rule out faults, demand and manual actions
Before treating the drop as a quality judgement, check the Manual actions and Security issues reports in Search Console, the Page indexing report for sudden rises in excluded pages, and your deploy log for releases in the same week. Look at robots.txt, canonical tags and noindex settings on the pages that fell. If positions held while impressions dropped, the cause is demand or result layout, not the update.
Step 3
Measure the loss page by page
In the Search Console Performance report, compare whole weeks after the rollout with the same number of weeks before it, export the Pages and Queries tables, and calculate the position change for each. Sort pages into small slips, large falls and steady pages, and group them by type such as guides, reviews or service pages. Concentrate your effort on large falls in groups that matter to the business.
Step 4
Study the pages that replaced yours
For your ten or twenty most valuable lost queries, search in a private window and record the top five results in a spreadsheet: who wrote each, what type of page it is, what original material it includes, how current it is and how quickly it answers the question. Look for features the winners share. Note any query where the kind of result has changed, such as guides replaced by product pages or forum threads.
Step 5
Review the losing pages honestly
Read each losing page as a stranger would and score it against Google's published self-assessment questions: does it offer original information or analysis, does it fully answer the question, would you trust it, and does it show first-hand experience? Ask someone outside the business to do the same for a sample. Be specific in your notes, for example "no prices, no photos of our own work, answer appears after 400 words".
Step 6
Improve, merge or retire weak content
For
pages worth keeping, make substantial changes that close the gaps you found: add first-hand detail, real
examples, clearer structure and a named author with relevant background. Merge overlapping pages into
one stronger page and 301 redirect the others to it. Remove or noindex pages
that help nobody. Avoid cosmetic edits such as new dates or reworded titles, which do not change how
useful a page is.
Step 7
Track recovery through the next core update
Keep a dated change log listing each page you improved and what you changed. Use the Algorithm Impact monitor to track recovery week by week, and compare improved pages against untouched ones from your original export. Expect gradual movement at best between updates, then review the whole list again after the next core update completes, because that is often when broad quality changes show most clearly.
The quality questions that matter most
Google publishes a list of questions site owners can use to assess their own content. You do not need to memorise them, but the ones below are where most sites that lose in core updates fall short. Rephrase them for your own pages and answer honestly:
| Question | Weak answer | Strong answer |
|---|---|---|
| Does the page add something original? | A summary of what other articles already say. | Your own data, photos, test results, case notes or local knowledge. |
| Does it fully answer the question? | Covers the basics and sends the reader elsewhere for the detail. | A reader could act on it without searching again. |
| Is it clear who is behind it? | No author, or a generic "Admin" byline. | A named author with a short, relevant background and an about page. |
| Would you trust it with money or health? | Claims without sources; outdated figures. | Sources cited, figures dated, limitations stated plainly. |
| Was it written for people or for rankings? | Keyword-stuffed headings, padding, topics chosen only for search volume. | Written because your audience needed it, in the words they use. |
| Is the answer easy to reach? | Long preamble, pop-ups and adverts before the answer. | The answer or summary near the top, detail below. |
Google also suggests asking "who, how and why" about content: who created it, how it was produced, and why it exists. If the honest answer to "why" is "to rank for this keyword", that page is a candidate for rewriting or removal. The E-E-A-T Guide goes deeper into the experience and trust signals behind these questions.
When waiting is the right call
Not every drop needs a response. Rushed changes can make a page worse, and a core update sometimes reshuffles results in ways that partly unwind later. Waiting and monitoring is reasonable when:
- the losses are small position changes on pages that remain on page one;
- the pages that replaced yours are no better than yours on any of the questions above;
- the drop affects a few queries rather than a section or the whole site;
- you have already reviewed the pages recently and they are strong.
Waiting is the wrong call when you lost large amounts of traffic across a whole section, when the winners are clearly more useful, or when you know parts of the site were built mainly to attract search traffic. In those cases, every month spent waiting is a month of improvement not made before the next update.
Changes that waste time or make things worse
- Changing publication dates. Updating a date without updating the content does not make it fresher in any useful sense, and readers notice.
- Mass rewriting with AI tools. Rephrasing weak pages at scale produces the same weak pages in different words, and publishing many pages mainly to gain rankings can fall under the scaled content abuse spam policy.
- Adding length for its own sake. There is no ideal word count. Padding pushes the answer further down and can make the page less useful.
- Deleting large sections in a panic. Pages that still earn traffic, links, leads or sales should be improved, not removed. Prune what helps nobody.
- Disavowing links as a reflex. A core update is not a link penalty. Disavowing links that were helping you can remove value.
- Buying links to compensate. Paid links intended to pass ranking credit break Google's spam policies and add a second problem to the first.
Platform notes
WordPress
WordPress sites often carry large numbers of low-value URLs that dilute the site as a whole: tag archives
with one post each, date archives, attachment pages and author archives on single-author blogs. Your
SEO plugin can set these to noindex. Check the Page indexing report afterwards to confirm
Google has processed the change.
Shopify
Collection pages filtered by tag and products reachable at several paths create near-duplicate pages. Shopify themes normally add canonical tags, but thin collection pages with no introduction and a handful of products are a common weak spot. Add genuinely useful buying guidance to the collections that matter, and combine collections that differ by one product.
Custom and static sites
Keep a machine-readable change log, for example a CSV with URL, date and a one-line description of
the change, committed alongside the site. It makes the before-and-after comparison in step 7 far easier.
Keep lastmod dates in your sitemap honest, updating them only when content has meaningfully
changed.
1. Why did my site drop in a core update when I did nothing wrong?
Because a core update changes how Google assesses content in general, not whether you broke a rule. Google has said plainly that a drop does not mean a page has done something wrong. It may simply be that other pages are now judged more useful for the same searches, or that the type of result searchers want has changed. That is why the recovery work focuses on comparing your pages with the new winners and genuinely improving them, rather than hunting for a single error.
2. How long should I wait after a core update before making changes?
Wait at least until the rollout is confirmed as complete on the Google Search Status Dashboard, and ideally another week or two so positions can settle. Changing pages mid-rollout makes it impossible to tell whether later movement came from the update or from your edits. Use the waiting time productively: export your Search Console data, list the pages that lost most, and study the pages that replaced yours, so you are ready to act as soon as things settle.
3. Will my rankings come back at the next core update?
They might, but nothing guarantees it. Google has said that improvements can be recognised over time, and that the clearest changes after broad quality work often appear when a later core update runs. Sites that make substantial, genuine improvements have a better chance than sites that make cosmetic edits or change nothing. Some pages never return to their old positions because the results have permanently shifted towards a different type of content, and that is worth accepting early.
4. Is a core update drop the same as a penalty?
No. A penalty in the usual sense is a manual action, which a human reviewer at Google applies for breaking the spam policies and which is reported in the Manual actions section of Search Console. A core update is automated and affects huge numbers of sites at once, most of which have broken no rules. There is no reconsideration request for a core update, and no message to tell you what to change. Recovery comes from making content more useful.
5. Should I delete old content after a core update?
Only content that helps nobody. Because Google's systems can weigh the site as a whole, a large volume of thin or outdated pages can hold back the good ones, so removing or merging genuinely weak material can help. But pages that still earn visits, links, enquiries or sales should be improved rather than deleted. Before removing anything, check its traffic over the past year and its backlinks, and redirect merged pages to the closest useful replacement.
6. Does updating the date on my articles help with core updates?
Not by itself. Changing the visible date without changing the content does not make a page more useful, and Google has advised against doing it to appear fresh. If you genuinely update an article with new information, corrected figures or better examples, it is reasonable to show an updated date and say what changed. For topics where current information matters, real updates can help readers and rankings. For evergreen topics, the date matters much less than the substance.
7. Can AI Overviews explain a drop that looks like a core update?
They can explain a drop in clicks, but not a drop in position. If your average position in Search Console stayed roughly the same while clicks fell, the cause is more likely a change in the results page, such as an AI Overview or other feature appearing above you, or a fall in searches. If your position itself fell sharply during a confirmed rollout, the update is a more likely cause. Check both measures before choosing which problem to work on.
The Algorithm Impact monitor lines up your daily ranking movements with tracked Google updates, helps separate update effects from technical faults, content decay and competitor gains, tracks your recovery and suggests actions suited to the type of update.
Open Algorithm Impact Fixes →