A falling site-total graph does not tell you how much of your website is affected. One important page can lose enough search traffic to pull the whole chart down while other pages remain stable. Equally, a modest overall decline can hide losses across several sections if another part of the site is growing.
Before changing content or technical settings, establish where the loss sits. We narrow the investigation by comparing pages, query groups and search conditions, then checking whether the affected URLs share a cause. The useful answer is not always simply “sitewide” or “one page”. It may be a page cluster, a particular topic or searches from one country.
Here is how to check whether you have a sitewide or page-specific ranking drop, and how to turn that finding into a proportionate response.
Establish What Actually Dropped
Start in the Search Console Performance report, even if a rank-tracker alert prompted the investigation. A tracked keyword gives you a reason to look; it does not establish the scope of the problem.
Review clicks, impressions and average position together. Each answers a different question:
- Clicks show whether fewer visits are arriving from Google Search.
- Impressions show whether your results are being seen less often.
- Average position helps you investigate changes in where those results appear.
Fewer clicks alone do not prove that rankings have fallen. If impressions remain steady, examine position alongside the titles, snippets and competing results that searchers see. Google notes that changes in how appealing results look can affect clicks without the same decline in impressions.
When clicks and impressions both fall, investigate visibility and search demand. A seasonal query can generate fewer impressions because fewer people are searching, rather than because your page has become less competitive.
Treat the site-wide average position as a starting signal, not a diagnosis. It combines different pages and queries. To understand whether an important ranking has weakened, compare the relevant page and query under consistent conditions.
Build a Fair Date Comparison
Find when the decline began, then compare the affected period with a similar earlier period. Keep the date ranges comparable and avoid drawing conclusions from an unusually strong day followed by an ordinary one.
Use the longer history as context. Search Console’s last 16 months view can help you spot a recurring seasonal pattern. If the same topic lost traffic around the same time in the previous year, investigate demand before assuming a new website fault.
Keep the search type consistent. A decline in image search should not automatically trigger changes intended to address web search rankings. Review those search types separately.
Record the filters you use, including country and device. A country filter matters when your business serves a particular market: an international traffic decline may not describe what happened to your commercially important searches in the Philippines. Likewise, a mobile-only change needs a different investigation from a decline visible across devices.
Your first working note should identify the affected dates, search type, filters and metrics. This makes later comparisons meaningful and stops the investigation shifting between incompatible views.
Separate Normal Movement From a Significant Decline
Google states: “Small fluctuations in position can happen at any time”. Its guidance also advises against radical changes when a page is already performing well.
The distinction in Google’s search-traffic debugging guide is useful here: a small shift among the top results is different from a large drop out of those results across many terms. Both can reduce clicks, but they do not justify the same response.
Look at persistence as well as severity. A brief movement that reverses is weaker evidence for intervention than a sustained deterioration across important queries. That does not mean ignoring a confirmed technical fault. An accidental indexing restriction deserves attention even while traffic data is still catching up.
Where movement is small and there is no corresponding fault, retain the comparison and monitor it. Rewriting a stable page immediately can remove useful content while making it harder to understand the original change.
Find Which Pages Account for the Loss
With the date comparison active, open the Pages table and order it by Clicks Difference so that the largest losses are easy to inspect. This shows which URLs contribute most to the decline, rather than which ones happen to have caught your attention.
Look beyond the first losing URL. Ask whether losses are concentrated in one page, spread through a recognisable section or visible across otherwise unrelated parts of the website.
A practical classification is:
- Page-specific: one URL accounts for the main loss while comparable pages remain broadly stable.
- Section-specific: affected URLs share a directory, template or content type.
- Broad: losses appear across several distinct sections and query groups.
- Query-specific: movement follows a search topic more closely than a website section.
These are working descriptions, not proven causes. Several affected pages may share a template fault, but they may also target the same seasonal topic. The next checks need to distinguish those explanations.
Keep some stable pages in your comparison. An unaffected page using the same template can help test whether a suspected template change explains the pattern. Stable pages in another section also help establish the boundary of the loss.
This is where the site-total graph can mislead most. If one previously high-traffic article accounts for most of the decline, the business impact may be substantial, but the evidence does not yet support a website-wide repair.
Check the Queries Behind the Affected Pages
Select an affected URL and examine its queries using the same date comparison. Determine whether the page has lost visibility across its main searches or only within one query group.
A page that declines across several relevant queries calls for a broader review of that URL than a page that loses one term while retaining others. In the latter case, inspect what changed for that particular search before rewriting the whole article or commercial content.
Then reverse the view: filter for an important affected query and inspect the pages associated with it. Check whether another URL on your site is now receiving visibility for that query. A decline against the original URL does not necessarily mean the website has lost all of that search presence.
Where the pattern follows a topic, compare search interest using Google Trends. Google recommends this to help distinguish a website-specific decline from wider changes in demand. Use the relevant region so that international interest does not obscure the market you serve.
Keep the conclusion specific. “Our advice articles about this topic lost impressions” directs a useful investigation. “Google has dropped the site” hides the very distinctions needed to choose the next action.
Investigate Shared Causes When Losses Are Broad
When losses extend across unrelated sections, inspect shared technical conditions before commissioning widespread content edits. Check Search Console’s Page indexing and Crawl stats reports for corresponding changes in detected issues.
Server availability and problems fetching robots.txt can affect Google’s ability to access the website. Other faults may affect only particular URLs or templates. Compare the affected pages with the errors reported rather than assuming that every warning explains the traffic decline.
Review the deployment window as well. Did the loss follow a release, URL migration or change to indexing settings? Timing helps identify what to inspect, but coincidence alone does not establish the cause. The suspected change should also explain which pages were affected and how Google’s access or indexing changed.
Check the Security Issues and Manual Actions reports for explicit findings. Those reports can identify problems that require a different response from content optimisation. Their absence, however, does not by itself explain the decline.
Our technical SEO work covers crawlability, site architecture and indexing, followed by prioritised implementation and validation. That is the relevant route when the evidence points to shared technical barriers, rather than an instruction to rewrite every losing page.
If technical checks do not explain a large, persistent position drop across a wide range of terms, broaden the review to the website’s content quality and usefulness. Google recommends assessing the whole website in that situation. A nearby algorithm update is context for that assessment, not proof that one particular change will reverse the loss.
Inspect the Boundary of a Page or Section Problem
For a limited page cluster, use URL Inspection on representative affected URLs. Compare them with a stable URL where possible. Check the indexing state and investigate unexpected differences, including whether Google is treating another URL as canonical.
Review what changed on those pages: content, titles, internal links, URL destinations or indexing instructions. The important question is whether the change fits the observed loss. A content edit cannot explain an unrelated server outage, while stronger copy will not remove an accidental noindex instruction.
A misplaced noindex can also produce a delayed decline because Google needs to crawl the page to discover it. Do not restrict your change review to the exact day the traffic graph turned down.
If several affected pages use the same template, inspect the shared output as well as individual content. Confirm whether the suspected problem appears across that group. This prevents a section-wide fault from being treated as a series of unrelated writing problems.
Where URLs have moved, include their replacements in the investigation. Google explains that rankings can fluctuate while it recrawls and reindexes moved content. Looking only at the old URLs can give an incomplete account of what happened after migration.
Make the Next Action Match the Evidence
Finish the investigation with a short, testable finding. Record the affected scope, the metrics that changed, the comparison conditions and the evidence for the suspected cause. Separate what you have confirmed from what still needs checking.
A useful finding might identify a loss concentrated in one page cluster, with indexing differences still under investigation. That is more actionable than declaring a sitewide ranking problem before checking the rest of the website.
Choose the response accordingly. Correct a confirmed technical fault, investigate a topic-level demand change or review content where sustained ranking losses support that work. Preserve the original comparison so you can assess what happens after the intervention without changing the measurement conditions.
If the pattern remains unresolved, our SEO audit brings crawl and index analysis together with content and internal-link review, then turns the findings into a prioritised action plan. Bring the affected URLs, comparison dates, relevant Search Console views and recent change history so the investigation starts with the observed problem.
The practical test is whether losses spread across independent sections or remain concentrated once you separate pages, queries, countries and devices. Establish that boundary first. It tells you whether to investigate the website’s shared foundations, a particular section or one page, and helps you avoid changing healthy parts of the site without evidence.
