Validating Search Console Bulk Export Partitions and Retries
When Search Console sends performance data to BigQuery, it writes each day's data into date-based partitions, which keep each date's rows together.…
After you change a consent banner or Google consent mode, compare Search Console and Google Analytics 4 (GA4) separately. Search Console counts activity on Google’s search results, including clicks and impressions. GA4 counts activity recorded by its tags. Changes to tag behavior or consent signals can therefore change GA4 sessions even when search clicks stay steady. GA4 groups recorded activity into sessions, each counted as one visit.
Search Console counts clicks whether or not your GA4 tag loads. If Search Console clicks stay steady while GA4 organic sessions fall, first check whether GA4 measures or attributes visits differently. This pattern does not prove that the banner caused the entire difference. The documented measurement mechanisms explain why the two sets of numbers can differ, but do not show that a banner update directly changes Search Console totals. See Google’s definitions of Search Console metrics and consent mode behavior.
Record when the banner or tag change reached each affected page, region, or visitor group. Note what changed. For example, check whether tags used to wait for consent but now send cookieless measurements, which are measurements sent without cookies, before consent. Also note whether the default consent setting changed or the banner began sending updates at a different time. The comparison boundary is the date the change reached each group; someone closing a deployment task marks a different date.
Search Console does not provide a consent-state dimension in its documented Performance report breakdowns. It groups data by dimensions such as date, query, page, country, and device, so deployment records help you interpret those patterns. Country or device can reveal where a change is concentrated, but neither identifies a visitor’s consent choice. See Google’s documentation on Search Console dimensions.
Open Performance → Search results. Compare the same property and search type, with matching filters for the periods before and after the change. When possible, use complete weeks or months of similar length to reduce day-of-week differences. First compare clicks and impressions. Then check click-through rate (CTR), the number of clicks divided by impressions, and average position, which summarizes where results appeared. Search Console supports date-range comparisons and filters.
Wait for reporting delays before interpreting the latest period. Search Console data is usually available after two to three days, and the newest data can still be preliminary. The standard report uses Pacific Time for daily dates. Before comparing daily totals, align those dates with GA4’s reporting time zone. Google explains these data freshness and discrepancy limits.
Use the chart’s totals for the whole property as your main check. Then inspect the query and page tables. Check the country and device tables too to find where numbers changed. The table rows do not always add up to the chart total. Search Console can hide rare queries to protect privacy, and each table is limited to 1,000 rows. How Search Console groups pages and queries can also affect totals. See Search Console’s data reporting details.
In GA4, open Reports → Acquisition → Traffic acquisition. Compare session-scoped dimensions, which describe a session, such as Session source / medium. Filter for google / organic, then add a landing-page filter or dimension to match the pages you checked in Search Console. Also compare sessions with relevant event counts, such as page_view or session_start, if your setup records those events. Google’s Traffic acquisition report documentation describes the report dimensions.
Keep session-based reports separate from user-based acquisition reports, which group data by user rather than session. GA4 measures the two scopes differently, so a source can have different values in each report. The linked Google organic search traffic report in GA4 shows Search Console and Analytics metrics by landing page, country, or device. The two products still count different things. Search Console data appears in the linked GA4 reports 48 hours after Search Console collects it, and those reports cover up to 16 months. See Google’s linked report details.
Read the combined pattern as a diagnostic clue:
| Pattern after the update | What to check next |
|---|---|
| Search Console clicks stay broadly stable while GA4 organic sessions fall | Check tag firing, consent updates, and GA4 source attribution before treating the change as lost search traffic. |
| Search Console clicks and impressions also fall | Inspect the affected queries, pages, countries, and devices, then investigate search performance and other site changes. The banner update alone does not establish the cause. |
Search Console clicks stay stable while GA4 shifts from google / organic to Direct or an unassigned category |
Check landing-page redirects, referrer handling, campaign parameters, and whether GA4 recorded the session and its source. |
| GA4 sessions or users move differently from event counts | Confirm the report’s metric scope and whether modeled data is included before interpreting the movement as a visit change. |
Use Tag Assistant, Google’s tool for checking tags, to see whether the expected default consent setting appears before tags run and whether it changes after a visitor makes a choice. Check that the tags behave as intended for the visitor’s resulting consent setting. Google’s consent debugging instructions describe these checks.
Check timing if GA4 activity changes when the update reaches a page or visitor group. Google warns that GA4 can miss a consent update on a transition page in later reporting. For example, the report might be missing session_start. An update sent just before a page reload can also fail if the browser cancels the request. If the update changed how URLs or click identifiers are handled, inspect landing URLs and redirects. Redirects need to keep relevant parameters so GA4 can attribute visits correctly.
A GA4 comparison can also span a change in modeled data. Behavioral modeling estimates activity for some visitors who did not consent to analytics storage. Google applies eligibility and data-quality requirements to these estimates. Check the report’s data-quality indicator to see whether GA4 includes estimated data, excludes it, or has no estimate available. Also record the reporting identity used for the comparison. Google lists prerequisites that include at least 1,000 daily events with analytics_storage='denied' for at least seven days and at least 1,000 daily users with granted analytics storage on at least seven of the prior 28 days. Meeting these thresholds alone does not guarantee eligibility. See GA4 behavioral modeling requirements.
Give Vroni a GitHub issue, bug report, spec, or rough idea. It reads the repo, plans the change, writes code, runs checks, and works toward a review-ready pull request.
Take a look at vroni.com