Library·SOP·SEO
Find the likely reasons for a drop in Google Search traffic and decide what to check or fix next.
Last reviewed: · Editorial review
Clarity, structure and references reviewed in this version. This is not a test in a live account.
Find the likely reasons for a drop in Google Search traffic and decide what to check or fix next.
This SOP covers:
This does not cover:
First confirm that the report describes the right website and that both periods have usable data. A mismatch or unsupported date range can look like a traffic loss.
Make one copy of the shared results workbook.
Open the Monthly SEO Results — Traffic, Leads & Sales workbook and choose Make a copy. Use your copy for this investigation.
Read Read Me first. The Worked Example uses fictional data to show how to write a finding and an action.
Enter the overall comparison on Search Console Summary. Enter important pages, queries, evidence, next steps, owners, and follow-up dates on Search Console Details.
Open the Search Console report that matches the website and search type.
In Google Search Console, choose a property that covers the site. A Domain property covers the domain and its subdomains. A URL-prefix property covers only the exact protocol, host, and path shown.
Open Performance → Search results. Use Web for ordinary web searches. Keep Image, Video, or another type only if that is where the reported change appeared.
On Monthly Summary, enter the Search Console property under Search Console property and the report under Search type. Search Console Summary links to these fields, so leave its linked property and search-type cells as they are.
Compare two complete periods of equal length around the suspected change.
For a suspected drop beginning May 1, a clear first comparison is Apr 3–Apr 30 versus May 1–May 28: both periods are 28 days and contain the same weekdays. Use the dates nearest your case, and choose dates Search Console shows as complete.
In Performance → Search results, choose More time ranges to open the Date range dialog. Open the Compare tab and select the Custom radio option.
Enter the later/current period in the first Start date and End date pair. Enter the earlier/previous period in the second pair after vs. For the example, put May 1–28 first and Apr 3–30 after vs.
Choose Apply. Before recording a metric, check that the two date ranges on the cards or table match the periods you entered and that each value is under current or previous as labeled. If Search Console reports End date precedes start date, check that the later period is in the first pair and the earlier period is after vs; then apply again. If the property lacks history for either range, use dates within its history or report current-period results only.
Use Daily to find when the change began. If you are checking a longer pattern, also review Weekly or Monthly so ordinary day-to-day variation does not dominate.
Search Console data may take a few days to appear. End the comparison at the latest complete date shown, rather than including preliminary days.
If searches are strongly seasonal and the property has enough history, run one separate comparison with the same calendar dates from the previous year. Record those exact dates and the result in Notes; keep the recent equal-length comparison as your main finding.
Record the dates and settings exactly as shown in Search Console.
On Monthly Summary, enter the current and previous period start/end dates, comparison method, and shared Search Console scope or filters. Search Console Summary links to those dates, the property, search type, and shared filters. Fill only its Time granularity field; leave the linked cells unchanged.
After applying the comparison, match the current and previous values to the dates shown in the report cards or table. Do not assume which side is current based on how you entered dates.
Confirm the property and report cover the dates. A displayed zero can be a real zero when the report covers that period. If coverage is incomplete or unclear, do not substitute zero: use dates within the available history or report the current period alone and explain the limit in the Change % note or Notes field.
A site total shows the size of a change. Pages and queries show whether the change is broad or concentrated.
Use the applied Search results comparison to locate the change and describe its size.
In Performance → Search results, turn on Clicks, Impressions, Average CTR, and Average position. Clicks are visits from Google results; impressions are appearances; CTR is clicks divided by impressions; average position summarizes the topmost result from your site.
At Daily granularity, look for the first sustained change rather than a one-day spike. Review Weekly or Monthly as needed to see whether the longer trend agrees.
Enter the displayed current and previous values for all four metrics on Search Console Summary. Let the workbook calculate the differences.
Find which important pages and searches account for the change.
Open the Pages table and review the displayed click changes. Add a Summary row to Search Console Details, then add Page rows for important URLs and the largest losses that could change what you do next. In Page URL, query, or summary label, identify each row clearly.
Select an affected page and open Queries to see which visible searches changed for that page. Record selected Query rows with their displayed current and previous values.
Include pages the site owner considers important, even if they are not the largest losses. A focused shortlist is enough; you do not need to copy every URL or query.
Use the metric pattern to choose a useful check. Several causes can produce similar charts, so record what fits and what remains unproven.
Use the metric pattern to choose what to check next; treat it as a clue, not a diagnosis.
Clicks down; impressions similar or up; CTR down: the result appeared about as often, but fewer people clicked. Review the page title and description shown in search, the affected searches, and changes to the other results on the page. Search Console alone cannot tell which one caused the change.
Clicks and impressions down; average position worsened: check the affected pages for visibility, indexing, and ranking changes. Start with the most important pages, then inspect the specific report or URL if evidence points there.
Impressions down; position and CTR broadly similar: fewer searches may be happening, or the searches bringing impressions may have changed. Compare the affected queries and, when useful, broader topic interest in Trends.
Write the displayed pattern separately from your explanation in Evidence / likely explanation on the matching detail row. If the pattern does not fit one of these cases, record what moved and what evidence you still need.
Check for report issues and site changes around the date the decline began.
Review Search Console Data anomalies for an issue affecting your report type and dates. For a broad drop, also check Search Console Manual Actions and Security issues.
Ask the site owner or developer whether the site address, important URLs, redirects, page templates or content, robots rules, or site availability changed around the first affected date.
If the owner confirms that an important URL was removed, ask whether the removal was intentional and which current page should serve the same visitor need.
On the affected Page row, record what changed and when. If an important URL was removed, note whether the removal was intentional and the current replacement URL—or that none was named. Put the agreed next check in Next action, with its Owner and Follow-up date. On the Summary row, record which report or site clues matched.
If several important pages may share an indexing issue, use the report-level indexing guide linked to this step.
Use this route when Search Console shows a Page indexing reason affecting multiple important URLs. For a concern about one URL, use the single-URL route in the next step.
Bring the exact reason, affected-page count, one representative URL, and the report update date into your investigation notes.
If one exact page needs an indexing or canonical check, use the single-page guide linked to this step.
Use this route for one URL when Search Console shows an indexing or canonical issue, or other evidence points to one. A traffic loss by itself is not an indexing clue. Use the report-level route above when the same indexing issue affects several important pages.
Bring the exact URL, Search Console property, and dates when its traffic changed.
If an important query lost impressions, check whether interest in the topic may also have changed.
Open Google Trends and compare the same query or a broader topic that has enough data, in the geography that matters to the site. Use the same search type and date ranges where available.
Trends reports relative interest from 0 to 100, not a count of searches. A low-volume zero or one spike does not prove that demand disappeared. If Trends shows too little data, record that it cannot answer this question.
On the matching Query row, note whether topic interest rose, fell, stayed broadly similar, or could not be compared. Keep that note separate from the Search Console impression count.
If the timing overlaps a published Google ranking update, record it as one clue.
Compare the start and completion dates in the Google Search Status Dashboard with the change dates you found.
Record the overlap on the Summary row as a possible clue, not proof that the update caused the decline. Do not recommend a before-and-after conclusion until at least one full week after a core update finishes.
The best first action protects an important visitor or business result and has a supported, practical next check.
Choose the next checks by page importance, size of change, strength of evidence, and who can act.
Use this order: first, a supported problem on an important page or visitor task; next, the largest loss when the evidence narrows a useful check; then, weak or low-impact changes as Monitor with a revisit date. Put the order in Monthly Summary → Actions to take → Priority. Keep the supporting facts, Evidence / likely explanation, Owner, and Follow-up date on the matching Search Console Details row; that tab has no Priority field.
If there is not enough evidence to recommend a fix, write Monitor, say what information is missing, and set a date to check again. Do not make a percentage or average-position cutoff the rule for acting.
Fill Evidence / likely explanation, Next action, Owner, and Follow-up date on each row that needs work. Choose an owner who can complete the next check, not an unspecified team.
If several pages or searches need deeper comparison and prioritization, use the page-improvement guide linked to this step.
Use this route when you need to compare several pages or searches and decide which deserves work first.
Bring the property, Search type, date ranges, affected URLs or queries, and owner-named important pages.
If the question is whether one page meets a searcher’s need, use the single-page review guide linked to this step.
Use this route to review the visible title, description, and page content for one affected page.
Bring the page URL, affected query, and evidence that led to this question.
After the site owner approves a WordPress page edit, use the page-update guide linked to this step.
Use this route only after the site owner decides to update a specific WordPress page. This investigation assigns a next step; it does not approve a site change.
Bring the approved page URL, its purpose for visitors, and the findings that support the change.
Write a short conclusion that a site owner can act on and that separates facts from possible reasons.
Start with what changed, when it changed, and which important pages or queries were most affected. Include the comparison dates and Search type when they help explain the result.
Say what you checked and what the evidence supports. For example: “Clicks fell while impressions stayed similar and CTR fell, so we will review the main search results and affected pages.”
Name what is still unknown and give the next action, owner, and follow-up date. Avoid saying an update, competitor, or page change caused the decline unless the evidence confirms it.
