SEO Ranking Drop: How to Diagnose Yours Before You Start Fixing Things

Organic traffic fell, and the six causes Google names all produce the same symptom. Most teams respond by doing one of three things that don’t help: panic-publishing content, blaming whichever Google update they read about last, or “fixing SEO” at random. An SEO ranking drop has six causes that Google names, and you can narrow yours to one of them today.

The order of the checks matters as much as the checks themselves, because you can spend a quarter rewriting content to fix what turns out to be a robots.txt file.

Google logs known issues from the last 3-16 months that might affect your Search Console data (opens in new tab) on its data anomalies page. If your date range overlaps an entry there, part of your decline is a reporting artifact and not an SEO ranking drop at all.

Nine separate entries from February through August 2026 carry a data issue. A logging error on August 13, 2026 decreased clicks and impressions on the Discover performance report, and a second one from August 13 to 17, 2026 decreased impressions on the Generative AI performance report in Search, which Google restored on August 21. Google logged the same class of Discover error on June 24, 2026 and again on May 21, 2026.

Earlier in the year, a logging error from April 16 to 27, 2026 stopped Search Console from reporting impressions and clicks for job listings and details. Two days of bulk data exports went missing for some properties between February 28 and March 1, 2026, and that data won’t be recovered. A separate long-running impressions logging error ran from May 13, 2025 all the way to April 27, 2026 before it was resolved.

Starting May 7, 2026, FAQ rich results stopped appearing in Google Search, and you’ll see a drop in impressions for FAQ as a result. That’s a real, measurable impression loss with no relationship whatsoever to your content quality, your links, or your site speed.

Two more things in the Performance report can make a fresh SEO ranking drop an illusion. The newest data can be preliminary (opens in new tab): it’s still being collected and might change in the next few hours, which is why Google draws it with dotted lines. The chart totals can also sometimes differ from the table totals, because charts aggregate at the property level and tables aggregate at the page level.

So before anyone convenes a war room:

“Penalty” isn’t Google’s word for algorithmic movement. Google reserves human enforcement for manual actions, and the Manual Actions report (opens in new tab) is binary: either there’s an action against your site or there isn’t. Checking it takes about thirty seconds and it’s the fastest cause in the entire diagnosis to rule in or out.

If the report says no issues detected, Google penalty recovery isn’t the project in front of you. That matters, because the work is completely different. Google penalty recovery is a fix-and-appeal process with a ticket and an email trail. Recovering from an algorithmic SEO ranking drop has neither.

Core updates are broad in nature (opens in new tab) and don’t target specific sites or individual web pages. A core update reshuffling results isn’t a punishment aimed at you. It’s a change in how the whole index is evaluated, and your position moved relative to everyone else’s.

Google has fifteen manual action types, plus a separate set of News and Discover policy violations. The list includes unnatural links to your site and unnatural links from your site, thin content with little or no added value, cloaking and sneaky redirects, hidden text and keyword stuffing, user-generated spam, the site reputation policy, structured data issues, cloaked images, AMP content mismatch, sneaky mobile redirects, back button hijacking, spammy free hosts, sites abused with third-party spam, and major spam problems. If none of those describe anything anyone at your company has done, you almost certainly don’t have a manual action.

Google’s sequence is short. Expand the description panel in the report, see which pages are affected, read the linked explanation of the issue type, and fix it on every affected page. Then confirm Google can actually reach those pages: they shouldn’t require a login, sit behind a paywall, or be blocked by robots.txt or a noindex directive.

Only then do you select Request Review. Google says a good reconsideration request does three things: it explains the exact quality issue on your site, describes the steps you took to fix it, and documents the outcome of those efforts. Google confirms receipt by email and emails again when the review completes, and asks that you don’t resubmit before you get a decision.

Most reconsideration reviews can take several days or weeks, although in some cases, such as link-related reconsideration requests, a review may take longer than usual. Google publishes nothing firmer, so treat any specific date you’re quoted as an estimate rather than a commitment.

Run the Security Issues report (opens in new tab) in the same sitting. It covers hacked malware, code injection, content injection, URL injection, deceptive pages, harmful downloads, and several more. Code injection means a hacker has injected malicious code that does things like redirect to a malicious site or run cryptocurrency mining software, and the serving effect is a dangerous warning in the browser navigation bar plus a warning interstitial page.

Security reviews take a few days to a few weeks. The Uncommon Downloads issue flags a download that Google Safe Browsing hasn’t seen before. It won’t prevent your page or site from appearing in Google Search results, but it will show a warning in the Chrome browser when a user initially requests to download unknown files.

Google sorts the causes of a drop in Search traffic (opens in new tab) into six categories. Build the investigation on that spine instead of an invented framework, and every finding maps to a report you can open.

Algorithmic update. Google says it’s always improving how it assesses content and updating its search ranking and serving algorithms accordingly. The only authoritative way to tell whether there was an update that week is the Google Search Status Dashboard (opens in new tab), which tracks Crawling, Indexing, Ranking, and Serving with start times and rollout durations in US/Pacific.

Google’s confirmed 2026 ranking updates through September are the February 2026 Discover update (February 5, 21 days and 17 hours), the March 2026 spam update (March 24, 19 hours and 30 minutes), the March 2026 core update (March 27, 12 days and 4 hours), the May 2026 core update (May 21, 11 days and 21 hours), the June 2026 spam update (June 24, 2 days and 1 hour), the August 2026 spam update (August 18, 2 days and 16 hours), and the September 2026 spam update (September 24, 13 days and 16 hours).

If your SEO rankings dropped on a date that doesn’t overlap one of those windows, you can rule out a named update as the trigger, though Google ships smaller core updates continuously and doesn’t log every one on the dashboard.

Technical issues. These are errors that can prevent Google from crawling, indexing, or serving your pages to users, such as server availability, robots.txt fetching, page not found, and others. The reports that isolate them are Page Indexing and Crawl Stats, where Google tells you to look for spikes in detected issues. The cause can be site-wide, like downtime, or page-wide, like a misplaced noindex tag.

Security issues. If your site is affected by a security threat, like malware or phishing, Google may alert users before they reach your site with warnings or interstitial pages, which may decrease Search traffic. That often reads as a click-through collapse rather than a ranking loss, with impressions holding while clicks fall.

Spam issues. If your site doesn’t comply with Google’s spam policies for web search, your content might rank lower in results or not appear in results at all. Google defines scaled content abuse (opens in new tab) as “when many pages are generated for the primary purpose of manipulating search rankings and not helping users.”

The site reputation policy applies where third-party content is published on a host site “mainly because of that host’s already-established ranking signals.” Google rewrote that section of its spam policies in August 2026, renaming it from site reputation abuse and adding a carve-out for the European Economic Area. Enforcement runs through automated systems and, separately, through human-reviewed manual actions.

Seasonality and changing interests. Sometimes changes in user behavior will change the demand for certain queries, either due to a new trend or to seasonality throughout the year. Google’s recommended cross-check is Google Trends, to establish whether the decline is yours alone or category-wide. If the whole category fell, you don’t have an SEO ranking drop, you have a demand curve.

Site moves and migrations. If you change the URLs of existing pages on your site, you may experience ranking fluctuations while Google recrawls and reindexes your site. This one is usually knowable in a single Slack message, because someone on your team knows whether URLs changed. If they did, start with our guide to migrating a site without losing SEO rather than with a content audit.

Google’s recommended method for working through all six starts in the Performance report. Use 16-month date ranges with period-over-period comparison, separate search type, track average position, and analyze the pattern at the page level. Then cross-check Google Trends to see whether the decline is site-specific or category-wide.

Google separates a small drop in position from a large one. A small drop, its example being position 2 to position 4, produces noticeable click loss while impressions stay roughly unchanged. A large drop means falling out of the top 10 to around position 29 across multiple terms. Impressions down and impressions flat are two different investigations, and starting on the wrong one costs you weeks.

Google illustrates those six causes with graph shapes. Before you read a single report in detail, look at the shape of the line:

The site-wide cliff sends you to the status dashboard to check dates and to the Security Issues and Manual Actions reports, because Google maps that one shape to all three. The gradual slide sends you to Crawl Stats and Page Indexing.

The recurring wave sends you to Google Trends and to last year’s data. The anomaly sends you back to the data anomalies page, which is why that check belongs first rather than last.

When traffic falls within days of a deploy, the cause is almost never content quality. It’s the crawl and index pipeline. This is also the category where a technical SEO audit pays for itself fastest, because the findings are binary rather than debatable.

A robots.txt file tells search engine crawlers which URLs they can access on your site. It’s used mainly to avoid overloading your site with requests, and it isn’t a mechanism for keeping a web page out of Google (opens in new tab). A disallowed page can still be indexed if other sites link to it. To keep a page out, block indexing with noindex or password-protect the page.

The classic version of this failure is a staging Disallow: / shipped to production. It can take days to register in the Performance report, and it surfaces first in Crawl Stats (opens in new tab) host status. A robots.txt outage has its own fallback, in three stages. For the first 12 hours, Google will stop crawling your site but will continue to request your robots.txt file, and from 12 hours to 30 days, Google will use the last successfully fetched robots.txt file.

After 30 days, if the site homepage is available, Google will act as if there’s no robots.txt file and crawl without restraints. A prolonged robots.txt outage doesn’t end in a frozen crawl, it ends in an unrestricted one.

A stray noindex is faster and more total. When Googlebot crawls the page and extracts the tag, Google will drop that page entirely from Google Search results (opens in new tab), regardless of whether other sites link to it. There’s no partial version of this outcome.

Depending on the importance of the page on the internet, it may take months for Googlebot to revisit a page. Removing a bad noindex doesn’t restore the page that afternoon, so use the URL Inspection tool to request recrawling rather than waiting.

Those two directives also interact badly. For a noindex rule to work at all, the page must not be blocked by a robots.txt file, and it has to be otherwise accessible to the crawler. If robots.txt blocks it, the crawler will never see the noindex rule, and the page can still appear in search results, for example if other pages link to it. That’s how a site ends up with pages it thinks are removed and pages it thinks are indexed, and is wrong about both.

The Page Indexing report (opens in new tab) is where you date the regression. A spike in “URL marked ‘noindex’,” “URL blocked by robots.txt,” “Redirect error,” or “Soft 404” gives you a date, and a date gives you a deploy. That turns an argument about SEO into a diff.

One more rare but real cause takes one click to rule out. The Removals tool (opens in new tab) can hide URLs from results for about six months. Blocking a URL there doesn’t prevent Google from crawling your page, only from showing it in Search results. An accidental removal request explains a single-URL disappearance that no other report accounts for.

Google draws a hard line between permanent and temporary redirects (opens in new tab). Permanent redirects show the new redirect target in search results, and the indexing pipeline uses the redirect as a signal that the redirect target should be canonical. Temporary redirects show the source page in search results, and the pipeline doesn’t take them as a canonical signal. Google recommends a permanent server-side redirect whenever possible.

A migration that ships 302s instead of 301s is therefore telling Google to keep the old URLs canonical while the new ones sit in Search Console reported as “Page with redirect.” Nothing looks broken. Traffic just never arrives.

The other classic migration mistake is blanket-redirecting every unmapped legacy URL to the homepage. When a site redirects any unknown URLs to its home page instead of returning 404 errors, that’s a soft 404, and Google said in 2011 that such cases can have negative effects on its understanding and indexing of your site (opens in new tab). Google also said then that it treated 410 (Gone) the same as 404 (Not found), so choosing between them isn’t worth a meeting.

Canonicals fail more quietly. Indicating a canonical preference is a hint, not a rule (opens in new tab), and Google may choose a different page as canonical than you do, for various reasons. Some duplicate content on a site is normal and isn’t a violation of Google’s spam policies, so duplication by itself isn’t the problem.

Google ranks those hints in order of how strongly they can influence canonicalization (opens in new tab). Redirects are a strong signal that the target of the redirect should become canonical. A rel="canonical" annotation is a strong signal that the specified URL should become canonical. Sitemap inclusion is only a weak signal that helps the URLs that are included in a sitemap become canonical.

Google names the mistakes directly. It tells you not to use the robots.txt file for canonicalization purposes, and not to use the URL removal tool for canonicalization, because it hides all versions of a URL from Search. It tells you not to specify a URL fragment as canonical, as Google generally doesn’t support URL fragments. And it tells you to include a rel="canonical" link on the canonical page itself, also known as a self-referential canonical.

For ecommerce specifically, faceted and filtered URLs plus a loose canonical strategy is how “Duplicate, Google chose different canonical than user” starts appearing in Page Indexing while your intended category page quietly stops earning impressions. Google’s canonicalization methods can stack and thus become more effective when combined: a redirect, a canonical annotation, and a sitemap entry that all agree is a much stronger instruction than any one of them alone.

International sites have a fourth failure mode with a distinctive fingerprint. If two pages don’t both point to each other, the hreflang tags will be ignored (opens in new tab). Each language version must list itself as well as all other language versions. You can’t specify the country code by itself, because Google doesn’t automatically derive the language from a country code.

A broken hreflang cluster doesn’t sink a site. It sends the wrong market’s page to the wrong market, so the symptom is a country-segmented drop and the diagnostic is the Performance report’s country filter.

Why did traffic fall after our redesign? Because a redesign is usually a migration, and with any significant change to a site (opens in new tab), you may experience ranking fluctuations while Google recrawls and reindexes your site.

For medium-sized websites, it can take a few weeks or more before Google starts showing new URLs instead of old ones. Google tells you to keep the redirects for as long as possible, generally at least 1 year, so signals have time to transfer. A post-redesign SEO ranking drop is usually several of the failures above compounding at once, which is why we handle it as SEO migration work rather than as a content problem.

Google processes JavaScript pages in three phases (opens in new tab): crawling, then rendering, then indexing. Googlebot queues pages for both crawling and rendering, and once Google’s resources allow, a headless Chromium renders the page and executes the JavaScript. Google Search won’t render JavaScript from blocked files or on blocked pages, so a robots.txt rule covering your JS bundle directory quietly empties your pages.

That’s the kind of SEO ranking drop a content audit won’t find. Compare the rendered HTML in URL Inspection against the source HTML, and if the content only exists in one of them, you have your answer. Single-page apps have a related trap: Google’s guidance is to avoid soft 404s in client-side routing by using a JavaScript redirect to a URL for which the server responds with a 404 HTTP status code, or by adding a robots noindex meta tag to error pages with JavaScript.

Server health is the other half. 5xx and 429 server errors (opens in new tab) prompt Google’s crawlers to temporarily slow down with crawling, and already-indexed URLs are preserved in the index, but eventually dropped. Google decreases crawl rate in proportion to the number of individual URLs that are returning a server error, and treats 429 as a server overload signal. Google also warns against using 401 and 403 to limit crawl rate, because the 4xx status codes, except 429, have no effect on crawl rate.

Crawl budget gets blamed for far more SEO ranking drops than it causes. Google aims its crawl budget advice (opens in new tab) at sites with a million or more unique pages changing weekly, sites with ten thousand or more unique pages changing daily, and sites with a large share of URLs sitting in “Discovered – currently not indexed.” Google calls those numbers a rough estimate, not exact thresholds, but most B2B and ecommerce sites are nowhere near them.

Crawling isn’t a ranking signal, and improving your crawl rate won’t necessarily lead to better positions in Google Search results (opens in new tab).

Our rankings look the same. Why is traffic still falling? Because clicks and impressions move for reasons that have nothing to do with position. Google’s own small-drop example is position 2 to position 4, where clicks fall noticeably while impressions stay roughly unchanged, which means a modest slide inside the top of page one can cost real revenue without registering as a ranking collapse.

SERP feature changes do the same thing from the other direction. The FAQ rich result retirement on May 7, 2026 removed a result format across the board, and Google says to expect an impression drop for FAQ because of it. Nothing about the affected sites changed. The surface they were appearing on stopped existing.

Average position also hides movement by design. In charts, it’s the average position of the topmost result from your entire site. In tables, it’s the average position in Search results for the specific URL or grouping dimension shown in the table row. A site-level average can look flat while the three pages that actually drive pipeline slid off page one, so segment by page before you conclude your SEO rankings didn’t drop.

Query mix is the third mechanism. Search Console reports Generative AI in Search as its own performance surface, which is where Google logged and later restored that impressions logging error from August 13 to 17, 2026, and Discover is reported separately again. If your reporting mixes surfaces and search types together, a shift in where impressions are being served reads as a decline. Segment first, and if generative surfaces are a growing share of your category, treat answer engine optimization as its own workstream rather than as a footnote to SEO.

The URL Inspection tool (opens in new tab) has a limit worth knowing before you spend a day in it. A “URL is on Google” result doesn’t guarantee that your page is appearing in Search results, and the tool can’t assess quality or security compliance, manual actions, or whether a page actually ranks.

Google’s position on page experience (opens in new tab) is conditional. There’s no single signal. Core Web Vitals are used by Google’s ranking systems, but Google Search always seeks to show the most relevant content, even if the page experience is sub-par. Getting good results in reports like Search Console’s Core Web Vitals report or third-party tools doesn’t guarantee that your pages will rank at the top of Google Search results.

The last page-experience-named ranking update Google logged was the Page experience update for desktop on February 22, 2022, which ran 9 days. Nothing page-experience-named has been logged on the status dashboard as a ranking update since.

The Core Web Vitals report’s data comes from the CrUX report (opens in new tab), which gathers anonymized metrics about performance times from actual users visiting your URL. URLs that have a similar user experience are grouped, and a group’s status means 75% of visits to all URLs in the group experienced the group status shown. URL groups without enough data are omitted entirely, and the whole report is historical field data rather than a live test.

The Core Web Vitals thresholds (opens in new tab) are LCP at or under 2.5 seconds, INP at or under 200 milliseconds, and CLS at or under 0.1, all measured at the 75th percentile and segmented across mobile and desktop. INP replaced FID as a Core Web Vital on March 12, 2024 (opens in new tab).

That gives you a clean test. Pull your Core Web Vitals history for the window in which traffic fell. If the field data hasn’t moved, speed is very unlikely to be the cause of your SEO ranking drop, though the report trails real user experience so a very recent regression can be real before it shows up. Slow pages are still worth fixing, but justify the project on the outcome it reliably produces.

That outcome is conversion, and it’s measurable on your own site rather than inferred from Google’s algorithm. We’ve written about what checkout speed does to revenue, and it’s the argument that actually survives a CFO’s questions. If the numbers support the work, Core Web Vitals optimization pays for itself whether or not it moves a single position.

The belief that Google favors recency sends teams off to republish two hundred dated posts after an SEO ranking drop. Google marks the idea that it prefers fresher content, so you’d better keep tweaking your page, as false (opens in new tab).

By Google’s account, trivial edits plus a bumped publication date buy nothing at all.

What Google does run is narrower. Its freshness systems (opens in new tab) are various query-deserves-freshness systems designed to show fresher content for queries where it would be expected. Google’s example is recent movie reviews outranking production articles. That’s query-scoped, not a site-wide decay penalty.

Google also puts this among the questions it tells you to ask about search-engine-first content (opens in new tab): “Are you changing the date of pages to make them seem fresh when the content has not substantially changed?”

So the shortcut isn’t neutral. It’s on a list of things Google names as a warning sign.

Content decay is still real as an outcome. Pages lose traffic when they stop being the best answer, because intent shifted, a competitor published something better, your product line changed, or the SERP added a feature. Update because the page is now wrong or incomplete, and treat it as SEO content strategy rather than as date maintenance.

Google said in 2011 (opens in new tab) that the fact that some URLs on your site no longer exist or return 404 errors doesn’t affect how your site’s other URLs, the ones that return 200 (Success) status codes, perform in its search results. It also said that if some URLs on your site 404, that fact alone doesn’t hurt you or count against you in Google’s search results.

The Crawl Stats report classifies “Not found (404)” as “Possibly good” among crawl responses, sitting apart from the bad responses like 5xx errors, unreachable robots.txt, and redirect errors.

Soft 404s can have negative effects on how Google understands and indexes your site. 404s on pages you care about are a symptom worth chasing: if some of the pages that 404 are pages you actually care about, Google says you should look into why it’s seeing 404 errors when it crawls them. And a 404 that should have been a 301 forfeits the redirect’s canonical signal, which is a lost-signal problem rather than a penalty. Chasing a link checker to zero isn’t how you reverse an SEO ranking drop.

Rollouts run in weeks. The March 2024 core update took 45 days on Google’s own dashboard, the August 2025 spam update took 26 days and 15 hours, and the November 2024 core update took 23 days and 13 hours. Google’s instruction is to wait at least a full week after a core update completes before analyzing your site, which means judging your position mid-rollout is judging noise.

Updates also don’t hunt tactics. Core updates are broad in nature rather than targeted at individual sites, and Google said of its March 2024 update that there’s nothing new or special that creators need to do for it as long as they’ve been making satisfying content meant for people.

Policies, on the other hand, do arrive on published dates, and those are worth tracking. Google announced scaled content abuse, the site reputation policy, and expired domain abuse (opens in new tab) on March 5, 2024, and set site reputation abuse enforcement to take effect starting May 5, 2024, a two-month notice period published in advance. The scaled content policy applies “no matter whether content is produced through automation, human efforts, or some combination of human and automated processes,” and Google said it’s fine to use an old domain name for a new, original site that’s designed to serve people first.

One thing genuinely is obsolete: the Helpful Content System as a separate system. Google lists it among its retired systems. In March 2024, it evolved and became part of Google’s core ranking systems, and Google confirmed in March 2024 that “there’s no longer one signal or system used to do this.”

Panda folded into core in 2015 and Penguin in 2016. Anyone diagnosing your SEO ranking drop as an HCU hit is naming a system Google folded into core ranking in March 2024.

How long recovery takes depends entirely on which of the six causes you’re in, and Google’s ranges span hours to several months. Google publishes no fixed recovery window for any of them.

Google algorithm update recovery is the slowest of the six and the least controllable. Some changes can take effect in a few days, but it could take several months for Google’s systems to learn and confirm that the site as a whole is now producing helpful, reliable, people-first content. If it’s been a few months and you still haven’t seen any effect, that could mean waiting until the next core update. You don’t necessarily have to wait for a major core update to see the effect of your improvements, though, because smaller core updates ship continuously.

Google says to avoid doing quick-fix changes and to focus instead on changes that make sense for your users and are sustainable in the long term. It also says deleting content is a last resort, and only to be considered if you think the content can’t be salvaged.

Google said its August 2024 core update (opens in new tab) also aims to better capture improvements that sites may have made. Google has separately logged its own faults on the record, marking “Ranking is experiencing an ongoing issue” twice, for 26 days starting October 5, 2023 and for 4 days and 11 hours starting August 15, 2024.

Google’s recovery timelines run from general SEO changes (opens in new tab) to core update recovery (opens in new tab). For SEO changes in general, some might take effect in a few hours, others could take several months, and you likely want to wait a few weeks to assess whether your work had beneficial effects.

A reporting glitch resolves on its own within days, a manual action review can take days or weeks once you’ve actually fixed the violation, and a core update recovery can outlast two quarters of planning. Committing to a recovery date before you’ve identified which cause your ranking drop has is committing to a number you can’t influence.

Waiting is not the same as doing nothing, but the difference between the two is discipline about what you touch. Record the baseline first: export your Performance data by query and by page before you change anything, because once you start editing you lose the ability to prove what the SEO ranking drop actually cost you. If your analytics can’t attribute organic sessions to revenue at all, fix that in parallel with GA4 setup and reporting, since an unmeasured recovery is indistinguishable from no recovery.

Then work in this order:

When a team tells us SEO is not working, the useful next question is which of Google’s six causes they’ve actually ruled out. Usually the answer is none of them, and the feeling that SEO is not working turns out to be several unrelated problems averaged into one number. And if the real question is why is my website not ranking at all, rather than why it slipped, start with URL Inspection and Page Indexing, because a page that was never indexed has a different fix than a page that lost position.

There’s no need to sign in to Search Console every day (opens in new tab). Google recommends monthly checks and checks after site changes. Meeting all of the requirements and best practices in Google Search Essentials (opens in new tab) doesn’t mean that Google will crawl, index, or serve a page’s content. And Google has a line worth quoting in any vendor conversation: “it doesn’t cost any money to appear in Google Search results, no matter what anyone tries to tell you.”

You can run every one of these checks yourself, today, with tools you already own. Confirm the data is real, rule out a manual action, place the decline against Google’s six causes, read the graph shape, then work the technical pipeline before you touch a word of content. By the end of that sequence, you’ll know which SEO ranking drop you actually have, and the next decision stops being a guess.

If you’d rather have someone run it with you, that’s what the Growth Audit is. A clear, no-obligation look at why you aren’t getting seen or converting, with a roadmap for how to fix it.

You built something worth finding. We make sure it gets found.