Currently Available: Need a skilled Software Developer for your next project?
Categories
SEO, GEO and AEO

Verifying Your Search Console Bulk Export Setup

Google Search Console’s bulk export sends Search performance data to BigQuery, Google’s cloud data warehouse. After setup, check Search Console’s latest export status and BigQuery’s ExportLog table. The status page reports the latest attempt; ExportLog records successful exports and the dates they cover. Together, these checks show whether new data is arriving, which table or date is missing, and whether a later export recovered it.

A missing log entry does not prove that an export failed. Google does not write unsuccessful attempts to ExportLog, and daily exports do not run at a guaranteed clock time. Check the date and table in the log, then compare them with the latest status and any error messages before deciding that the export is broken.

Confirm new data in BigQuery

In BigQuery, a dataset is a container for tables. Check that the destination dataset has the two performance tables, searchdata_site_impression and searchdata_url_impression, and ExportLog. Then inspect recent successful exports:

SELECT
  namespace,
  data_date,
  epoch_version,
  publish_time
FROM <code>PROJECT_ID.DATASET_ID.ExportLog
ORDER BY publish_time DESC
LIMIT 50;

Replace PROJECT_ID and DATASET_ID with the project and dataset that Search Console uses. In each row, namespace identifies the exported table, data_date identifies the day of Search performance data, and publish_time shows when the export completed. Google records these dates and times in Pacific Time. Check both table names: Google may export the two tables separately, so a recent entry for one does not confirm that the other is current. The table reference describes the log fields and export tables.

Look for a recent data_date for each table, and use publish_time to see when each successful export arrived. Do not treat the newest date as missing simply because it has not appeared at its usual time. Google says exports run daily but their timing varies. After initial setup, the first export can take up to 48 hours; it covers the day of export and does not include earlier historical data. The setup instructions explain that initial delay and the lack of backfill.

Find dates with no successful export

To check for gaps, compare expected dates with successful entries in ExportLog. This query checks each table for the 14 days ending yesterday in Pacific Time:

WITH expected_dates AS (
  SELECT data_date
  FROM UNNEST(
    GENERATE_DATE_ARRAY(
      DATE_SUB(CURRENT_DATE('America/Los_Angeles'), INTERVAL 14 DAY),
      DATE_SUB(CURRENT_DATE('America/Los_Angeles'), INTERVAL 1 DAY)
    )
  ) AS data_date
),
expected_tables AS (
  SELECT namespace
  FROM UNNEST([
    'searchdata_site_impression',
    'searchdata_url_impression'
  ]) AS namespace
),
successful_exports AS (
  SELECT DISTINCT namespace, data_date
  FROM <code>PROJECT_ID.DATASET_ID.ExportLog
  WHERE agenda = 'SEARCHDATA'
)
SELECT
  t.namespace,
  d.data_date
FROM expected_tables AS t
CROSS JOIN expected_dates AS d
LEFT JOIN successful_exports AS s
  ON s.namespace = t.namespace
 AND s.data_date = d.data_date
WHERE s.data_date IS NULL
ORDER BY d.data_date DESC, t.namespace;

Replace the project and dataset identifiers before running the query. Set the date range to start no earlier than the day the export was configured, because Search Console does not backfill earlier data. Each returned row means the log has no successful export recorded for that table and date. It does not explain why: the export could still be pending, or an attempt could have failed. Review recent dates again after the scheduled export has had time to run, and check Search Console’s status before treating a gap as an incident.

Check status and investigate a gap

Open the Search Console property’s Settings → Bulk data export page. It shows the latest export attempt and details for reported problems, but it does not provide a day-by-day attempt history. Search Console also emails property owners and full users about export errors and their resolution. Transient connection errors do not necessarily trigger an email. Google’s monitoring guidance explains these limits and recommends using Cloud Logs Explorer to investigate unclear errors. Cloud Logs Explorer is Google Cloud’s tool for searching logs. Search there for search-console-data-export.

Treat a recent missing date as pending review when export timing could explain it and the status page shows no persistent error. Investigate a non-transient error, an error notification, or repeated gaps for dates that have had time to export. Google lists issues such as quota, billing, and permissions. The list also includes dataset location and an invalid dataset ID. It includes schema mismatch and Deadline exceeded as well. The status page and Cloud Logs Explorer help identify the reported cause; correct the configuration or underlying issue that the error names.

Search Console’s Test report checks settings it can validate, but a passing test neither starts an export nor checks every possible failure, including quota problems. After correcting an issue, check the next scheduled export rather than treating the test as proof that data has arrived.

Confirm that a retry recovered missing data

For a non-transient export error, Search Console retries on a later scheduled export and includes previously unexported data. To confirm recovery, look for a new successful ExportLog row with the missing data_date. Check both table namespaces because the two tables can have separate export records.

Google says it keeps failed-day data for about a week before dropping that day’s data. After about a month of failed export attempts, Search Console stops the bulk export. These are approximate time windows, so investigate persistent gaps promptly rather than relying on the final day of either window. Since neither ExportLog nor the status page gives a complete history of failed attempts, use the latest status, notifications, Cloud Logs Explorer, and later successful log entries together.

What I'm building

Delegate tasks. Get software.

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

Email updates

Usually a new article and a few links I found interesting.

No spam. Unsubscribe with one click.

Leave a Reply

Your email address will not be published. Required fields are marked *