How to Diagnose and Fix a Sudden Traffic Drop

Google’s guide to debugging drops in Search traffic opens with a set of sketches that pair graph shapes with causes: a large drop from an algorithmic update or a site-wide security or spam issue, seasonality, a technical issue across the site or changing interests, and a reporting glitch. The shape of the line in Search Console is the first clue to the cause. The diagnosis works in a fixed order: confirm the drop is real, read its shape, line up its date, rule out technical problems, and only then decide whether an update did it.

Step one: confirm the drop is real, and that it is search

Start in Search Console’s Performance report, not in a rank tracker or an analytics dashboard. Search Console’s figures come from Google Search itself. If your analytics tool shows a cliff while Search Console holds steady, look at measurement first: a removed or broken tag, a consent or tag manager change, a deploy that dropped the script. If the tag was the cause, fixing it makes the “drop” disappear.

Two more checks belong in step one:

Google’s side of the data. Google says to check Search Console’s data anomalies page, because a drop can come from a change in data processing or a logging error.
Which search type fell. The Search type filter separates web search from Google Images and the Video and News tabs. Google’s AI features documentation says sites appearing in AI features such as AI Overviews and AI Mode are counted in the Performance report within the “Web” search type. A separate Generative AI performance report shows their impressions in AI Overviews and AI Mode.

A drop you see only in a third-party rank tracker, on a single keyword, for a single day, is not yet a drop.

Step two: read the shape

Google’s guide sets out the filters that turn one line into a diagnosis:

Look at 16 months. Set the date range to the last 16 months, so a yearly dip shows up as the repeat it is.
Compare periods. Compare the last three months with the previous period, or year over year, and click through the queries, pages, countries, devices and search appearance tabs to find where the change sits.
Sort the pages. In the Pages table, sort by Clicks Difference to see which pages lost the most. The key question is scope: the whole site, a group of pages, or one important page.
Compare clicks with impressions. If impressions held and clicks fell, Google points to the title link and snippet, or to other sites having a more appealing rich result. If both fell, work through the causes in the next steps.

Google also separates two kinds of ranking loss:

A small drop in position, such as position 2 to 4. It can show as a noticeable drop in traffic without a big change in impressions. Google says small fluctuations can happen at any time and recommends avoiding radical changes if the page is already performing well.
A large drop in position, out of the top results for a wide range of terms. For this pattern, Google asks you to self-assess the whole website, not just individual pages.

Step three: line up the date

Pin the day the drop started and compare it with three timelines:

Google’s. Google says it posts notable ranking updates on its ranking update history in the Search Status Dashboard.
Yours. Your own change log: deploys, migrations, redesigns, CMS and plugin updates, template changes, robots.txt and meta edits. A drop that starts the day after your release points at your release.
Demand’s. Filter the Performance report to your highest-traffic queries one at a time and check them in Google Trends. Google’s guide recommends this to tell whether the drop is yours alone or happening across the web.

A date that matches an update is a strong signal, not a verdict. Rule out a same-week change of your own first.

Step four: rule out technical, security and spam problems

Before concluding that an update did it, walk through the problems Google names in its guide:

Technical issues. Check the Crawl Stats report and the Page indexing report for a spike that matches the drop date: server errors, robots.txt fetch failures, pages not found. Google adds a detail that changes how a drop looks. A misplaced noindex depends on Google crawling each page to see it, so it produces a slower drop than a site outage.
Security issues. If Google detects malware or phishing, it may warn users before they reach your site. Check the Security Issues report.
Spam issues. Check the Manual Actions report in Search Console. A listed action is its own cause with its own process: fix the violation on every affected page, then file a reconsideration request.

If the Page indexing report shows a jump in “Crawled – currently not indexed”, read it as Google defines it: the page was crawled and not indexed, and Google gives no reason. It says where to look. It does not name the cause.

Step five: reconstruct recent changes

When the drop is real, Google shows no data anomaly, and no manual action is listed, the change log is where to look. Put two lists side by side: the URLs that lost clicks in the Performance report, and every URL change, redirect, canonical edit, noindex, robots.txt edit and deletion in the days before the drop. These are the changes that leave a signature.

A redirect to the wrong kind of page. A cleanup that sends an informational article to a commercial category page fires correctly and can still lose the ranking, because the destination does not answer the query. Google’s site move guide warns that redirecting many URLs to one irrelevant destination might be treated as a soft 404.
A canonical changed by a template. If page A ranked and a template change points its canonical at page B, Google may index B in A’s place. The page still loads normally for a visitor. The change is visible only in the head of the document or the response headers.
A noindex shipped in a template. Every page on that template drops out as Google recrawls it, so the loss spreads across the template’s URLs as recrawling proceeds.
A robots.txt block. Blocking a section stops Google from crawling it. It does not guarantee removal: the Page indexing help notes that a blocked page can still be indexed from links pointing at it, without its content.
Deleted pages. They return 404, and Google keeps trying them for a while before crawling them less and less often. The links that pointed at them now lead nowhere.
A URL pattern change without mapping. A platform or template migration that changes URLs without one-to-one redirects can produce the same losses at scale.

Recovery is to undo the change: restore the page at its original URL, remove the wrong redirect or canonical, then ask for a recrawl with URL Inspection. Google says a request does not guarantee indexing, and recovery depends on recrawling. What the fix does is put back the page and signals that were ranking.

Step six: if an update is the cause

A matching date, a site-wide pattern and no technical, security, spam or recent-change explanation point to an algorithmic cause. Google’s guide says there might not be anything fundamentally wrong with your content, and that search results change as the web changes.

For a small drop in position, avoid radical changes.
For a large drop, assess the whole site against Google’s guidance on helpful, reliable, people-first content, and improve or remove what falls short.

On timing, Google’s core updates documentation says “some changes can take effect in a few days,” while it could take several months for its systems to confirm that a site as a whole now produces helpful content. It adds: “If it’s been a few months and you still haven’t seen any effect, that could mean waiting until the next core update.” Google also says there is no guarantee that changes will produce a noticeable effect.

Step seven: say what you found before you change anything

Give stakeholders three things before work starts: the category, the fix path and a realistic timeline. For a tracking fault or a shipped noindex, the timeline is shorter once the fix is in. For an algorithmic cause, the timeline to share is the one in Google’s core updates documentation: days for some changes, and up to several months before its systems confirm a site-wide improvement. Pressure to “do something” can produce overlapping changes that make recovery harder to read.

Three problems that look like a traffic drop

Pages that never made it into the index. They show no drop because they never had traffic. The Page indexing report lists each URL’s reason for not being indexed. Start there.
Traffic back, conversions not. If Search Console clicks recovered but leads or sales did not, the problem sits after the click. Check that conversion tracking still fires, and look at what changed on the landing pages.
Fear of spam links pointed at the site. Check the Manual Actions report first. Google’s disavow guidance calls the tool an advanced feature that most sites will not need.

Prevent the next one

Self-inflicted drops can come from changes that ship without anyone checking their search effect.

Keep a change log of every URL change, redirect, canonical edit, noindex and deletion, with dates. With one, the next diagnosis goes faster.
Add a review step for those changes before they go live.
Run a full crawl after an incident, so a second change made in the same window does not stay hidden behind the first.

Frequently asked questions

Is my drop real, or a tracking problem?

Compare Search Console’s Performance report with your analytics. Search Console’s figures come from Google Search, not from your tag. If analytics fell and Search Console did not, check the tag first. Also check Search Console’s data anomalies page for a problem on Google’s side.

Did a Google update cause the drop?

Compare the start date with the ranking update history on the Search Status Dashboard, where Google posts notable ranking updates. Treat a match as a signal, not proof, until you have ruled out technical, security and spam issues and any change you shipped the same week.

My rankings dropped overnight after a site change. What should I check?

Line up the URLs that lost clicks with the change log. Check redirects to the wrong kind of page, canonicals changed by a template, a noindex in a template, robots.txt edits and deleted pages. Undo the change, restore the page at its original URL and request a recrawl.

How long does recovery take?

It depends on the cause. A fixed tag shows at once, and a restored page can return as Google recrawls it. For algorithmic drops, Google says “some changes can take effect in a few days,” that confirming a site-wide improvement can take several months, and that a few months with no effect could mean waiting for the next core update.