Hit by a Google Update? How to Investigate a Traffic Drop

Step 1 Confirm which traffic has fallen

Write down the metric, report, property and dates behind the concern. “Traffic is down” might mean fewer GA4 sessions, fewer Google Search clicks, fewer Discover clicks or fewer enquiries. Those are different problems.

In GA4, use Reports, Acquisition, Traffic acquisition and Session default channel group to compare visits. Separate Organic Search from Direct, Referral and other channels. Add Session source/medium when you need to isolate google / organic from other search engines.

GA4 Traffic acquisition guidance

Then check Search Console for Google clicks and impressions over comparable dates. Its clicks and GA4 sessions need not match exactly because the tools measure different things. Look for the pattern rather than expecting identical totals.

A total traffic decline is not automatically an SEO decline

Consider this fictional example, using equal reporting periods:

Total sessions fell 40%, but the recorded Organic Search sessions stayed level. The immediate investigation belongs with Direct traffic and measurement. This example does not establish why Direct fell; it shows why the total alone cannot support a claim that Google reduced search traffic.

Check completed enquiries, bookings or sales separately. If recorded search visits are stable but enquiries fall, test the enquiry process and assess lead quality before assuming a ranking change.

Step 2 Check the measurement before changing the site

Confirm you are viewing the intended website property with the same report filters. Compare complete periods, avoiding an incomplete latest day or a different number of weekdays. Search Console’s newest data can be preliminary; unavailable values can also export as zero.

Search Console report setup and data cautions

• Ask whoever manages analytics to check recent tag, consent-banner, plugin and channel-group changes. Confirm that page visits and completed enquiry or purchase events still record as intended.
• Check whether the concern comes from the original report or a spreadsheet, dashboard or export. A broken connector, missing export days or changed calculation can create a drop in the summary.
• Open Google’s Search Console data anomalies page. Match the affected dates, report and metric; a Discover logging error does not explain every decline in Web Search.
• Save the original comparison and note when you collected it. If Google later restores missing data, refresh the analysis rather than continuing to use an earlier export.

Known Search Console data anomalies

The dated appendix below includes reporting examples from 2026. Keep a data-quality note beside any comparison that crosses an affected period.

Step 3 Identify the update type and its scope

Use Google’s ranking history to confirm the update’s name, start and completion dates. The newsletter publication date is not the rollout date. Add your own site changes and first observed decline to the same record.

Google’s official ranking update history

Core updates

Core updates are broad changes to Google’s ranking systems, rather than actions directed at a particular website. A fall near a core update is not, by itself, evidence of a penalty or a specific fault.

Google recommends waiting at least a full week after a core rollout completes before analysing its effect in Search Console, then comparing with a period before it began. Use this for ranking assessment; fix a confirmed outage or broken tracking immediately.

Google’s core update guidance

Spam updates and manual actions

Spam updates improve Google’s automated spam detection. Where a sustained change overlaps one, review the spam policies and practices actually present on your site. Do not infer a violation solely from the date.

A manual action is a separate finding by a human reviewer at Google. Check Search Console’s Manual Actions report. If an action exists, follow its specific remediation and review process. An empty report does not rule out automated spam detection.

Google’s spam update guidance

Search Console Manual Actions report

Discover updates

Discover is an interest-based feed rather than a list of results for a typed query. Its traffic is less predictable than keyword-driven search. Assess its Performance report separately, including affected pages and countries. The report may be absent when its impression threshold is not met.

A Discover decline can reflect changing interests or distribution rather than a loss of ordinary Web Search rankings. The February 2026 rollout initially targeted English-language users in the US; do not assume a worldwide effect from the announcement alone.

Google’s Discover guidance

Step 4 Build a useful before and after comparison

Open Search Console’s Search results Performance report. Start with clicks and impressions, using the same search type in both periods. Keep Web: text-based, Web: multimodal, Images, Video and News views clearly labelled instead of mixing them.

Use the Date filter’s comparison options to choose matching periods. For a recent decline, equal 28-day windows can help balance weekdays. Review a longer view as well, including the same season last year when data is available.

For a core update comparison, place the baseline before the rollout and the assessment period after the completed rollout and recommended waiting period. Record the exact dates. A default “previous period” can accidentally include rollout days.

Search Console filtering and comparisons

Compare the segments that can reveal a pattern

• Pages: identify the URLs with the largest absolute click losses. Group them by service, topic or shared template. Keep some stable pages as a comparison.
• Queries: check which searches declined for the affected pages. Separate your brand queries from service or informational queries, and look for a change in the questions attracting visits.
• Devices: compare mobile, desktop and tablet. A concentrated mobile decline merits a check of mobile content, navigation and any device-specific fault.
• Countries: compare the markets you serve. A decline in one country calls for investigation of local demand, language targeting and access, rather than an assumption of a site-wide problem.
• Search appearance: examine relevant feature views where available. A retired feature or changed reporting category can affect its row without establishing a general ranking loss.

Click a page row to filter the report to that URL, then inspect its Queries, Devices and Countries tabs. Record each filter and clear it before assessing the next site-wide view. Query and URL filters can change totals because some data is omitted or truncated.

Search Console dimensions and grouping

Prioritise both business importance and absolute losses. A page falling from ten clicks to zero loses fewer visits than a service page falling from 500 to 400, even though its percentage decline is larger. With a zero baseline, a percentage change is undefined; label it rather than inventing one.

Step 5 Interpret the pattern and test alternatives

If impressions and clicks both fall, investigate demand, visibility and eligibility. If clicks fall while impressions are broadly steady, inspect query-level position and search presentation. The site-wide average can conceal a change in which queries your pages appear for.

Review actual search results for representative queries in the relevant market and device context. Note changes in competing pages, titles, snippets and result layouts. A current result is a snapshot, not proof of what every customer saw earlier.

Do not infer that AI answers caused the loss merely because clicks fell. Establish what the reports and observed search presentation can support. Keep a clear distinction between evidence and the explanation you are testing.

Google’s diagnostic guide to search traffic drops

Check seasonal demand

Use Google Trends for important lost queries or topics, selecting the relevant country and a period that shows recurring patterns. Compare with last year’s Search Console data and your own business cycle, including holidays, campaigns and changes in availability.

Trends reports relative interest, normalised to a 0–100 scale, rather than exact monthly searches. Low-volume terms can show zero. Treat a matching demand decline as supporting context, not conclusive proof of the cause.

Google Trends data and limits

Step 6 Check access security and indexation

Review Search Console’s Page indexing and Crawl stats reports for changes around the decline. Inspect important affected URLs and a stable comparison page. A shared template or directory can help narrow the investigation.

Use URL Inspection to compare Google’s stored result with a live test. Check crawl permission, page fetch, indexing permission and the expected canonical. Inspect the returned content where available; the live test cannot guarantee indexing or predict Google’s chosen canonical.

Google’s URL Inspection instructions

• Open the affected public pages and check for errors, unintended redirects, missing content or password protection. Check whether important URLs changed during a migration.
• Review robots.txt, noindex instructions and WordPress Search Engine Visibility settings. Confirm that the intended pages and required resources remain available.
• Check hosting and Cloudflare logs for relevant blocks, challenges or server failures. A page working in your own browser does not prove that a crawler receives the same content.
• Open Search Console’s Security Issues and Manual Actions reports. Investigate malware, injected pages or warnings promptly; keep this work separate from an ordinary content-quality review.

Save the settings before a correction, record its time and retest the same URLs afterwards. If it causes a new problem, restore the earlier configuration and verify that the rollback actually takes effect.

A clean technical test narrows the possibilities. It does not prove that an update caused the decline or that the content is already the best answer available.

Step 7 Improve the pages where evidence supports it

When a sustained visibility decline remains after the other checks, assess affected pages and the wider site. Compare the reader’s task with what the page delivers. Look for missing answers, outdated claims, weak original input or instructions that are difficult to follow.

Ask someone who understands the subject to review the evidence and usefulness. Accurate author information and sources help readers assess trust; adding a biography, FAQs or schema is not a guaranteed recovery technique.

Google’s helpful content self-assessment

Make changes that solve a documented weakness. Preserve useful pages and successful elements. Treat deletion as a last resort for content that cannot be made worthwhile; do not remove whole sections merely because they lost traffic.

Keep an action log with the page, finding, change, owner and date. Avoid simultaneous site-wide rewrites, new redirects and security changes that leave you unable to tell what changed the result.

How long can improvement take

Google says some changes may show an effect within days, while broader reassessment can take months. Improvements can be recognised between major core updates; waiting for the next announced rollout is not always necessary. No change guarantees a return to the earlier position.

For spam-related effects, compliance may need to be recognised over months. If a specific manual action is present, use its review process rather than treating it as routine algorithm volatility.

Keep the diagnosis reviewable

Write a short conclusion containing what you know, what remains uncertain and what you will test next. For example, using fictional findings:

“Google Web Search clicks declined 30% across matched 28-day periods. Most lost clicks came from three service guides, concentrated in mobile searches in our main market. The dates overlap an announced update, but the cause remains unconfirmed. We will verify mobile crawler access, inspect the affected queries and compare seasonal demand before deciding on content changes.”

Retain the exports and filters supporting that statement. Monitor the affected pages, stable comparison pages and business results using the same scope. Revisit the working explanation when new evidence contradicts it.

How to Measure SEO Success When Search Uses AI

A repeatable traffic drop checklist

• Confirm the report, metric, affected channel and first decline date.
• Validate tracking, filters, exports and reporting anomalies.
• Record the update type, scope and official rollout dates.
• Compare matched periods by page, query, device, country and search type.
• Check demand, security, access and indexation before changing content.
• Make evidence-based improvements, retain a rollback record and retest.
• Monitor consistent measures and state the limits of the diagnosis.

The useful outcome is a better-supported explanation and a focused action plan. A nearby update date or a promise of rapid recovery does not provide either.

Web appendix 2026 announced update timeline

Checked on 12 October 2026. These are the seven announced updates that began in 2026 and appeared in Google’s ranking history at that date. Dates follow the dashboard’s US/Pacific calendar. They provide context, not a diagnosis of any individual website.

The February Discover announcement initially specified English-language users in the US. Google also makes smaller unannounced ranking changes, so this table is not a complete record of every adjustment to Search.

Live ranking history and individual rollout details

Web appendix reporting checks for 2026

Check the live anomalies page before interpreting a dip. The following notes were verified on 12 October 2026. Match the metric and reporting surface, not just the date.

Google’s Search Console data anomalies record

An applicable reporting note can explain a data break without a loss of real visits. It does not rule out a separate performance problem. Annotate affected comparisons and check independent evidence.

Sources and further reading

Google — Diagnosing search traffic drops

Google — Ranking update history

Google — Search Console data anomalies

Google — Core update guidance

Google — Spam update guidance

Google — Discover guidance

Google — Performance report setup

Google — Filtering and comparisons

Google — Dimensions and grouping

Google — URL Inspection

Google — Manual actions

Google — Trends data

Google — Helpful content assessment

Google Analytics — Traffic acquisition