{"id":12675,"date":"2015-08-19T00:00:00","date_gmt":"2015-08-19T00:00:00","guid":{"rendered":"https:\/\/e-dialog.group\/blog\/data-sampling-problem-not-with-custom-tables-in-google-analytics\/"},"modified":"2026-02-26T13:37:57","modified_gmt":"2026-02-26T13:37:57","slug":"data-sampling-problem-not-with-custom-tables-in-google-analytics","status":"publish","type":"post","link":"https:\/\/e-dialog.group\/en\/blog\/data-sampling-problem-not-with-custom-tables-in-google-analytics\/","title":{"rendered":"Data Sampling Problem Not With Custom Tables In Google Analytics"},"content":{"rendered":"

\"\"The image above shows what appears to be a completely normal report: namely thisTarget completions<\/strong>forall goals<\/strong>in our e-dialog profile inYear-on-year comparison<\/strong>2014 to 2013 –so far so good<\/em>. Additionally we havetwo segments<\/strong>applied to the overall data to find out whether there were more target completions in Austria or Germany –so far so good<\/em>.<\/p>\n

With this report you can see wonderfully and at first glance that overall there were more visitors and more purchases from Austria than from Germany – this information will be incorporated into future company decisions.But did you also discover and read the information in the first line of the report?<\/em>\"data_sampling\"<\/p>\n

Data sampling<\/h2>\n

Atsoooo<\/em>large amounts of data(4,381,962 sessions)<\/em>and the additional use of a year-on-year comparison as well(two at the same time)<\/em>This analysis is based on segmentsno longer at 100%<\/u>of the data butjust more<\/strong>on a tight note31%.<\/strong>In order to access standard and custom reports more quickly, Google Analytics samples large amounts of data that would require too much computing power.<\/p>\n

Sampling means that not all of the data is used for the analysis but only a part of it – in this case a third. It will be thatSampling procedure<\/strong>applied.<\/p>\n

Sampling is widely used in statistical analysis because analyzing a subset produces similar results to analyzing all data. Additionally, sampling speeds up the processing of reports with large amounts of data, as they can be processed very slowly. This not only increases the loading time of the reports but also does not put too much strain on Google’s computing power.<\/p>\n

Data sampling is done by Google Analyticsautomatically<\/strong>, if for a reportmore than 500,000 sessions<\/strong>be recorded (seeGoogle Analytics Data Limits Documentation<\/a>). For reports that visualize visitor flow, sampling occurs as early as 100,000 sessions.But what if an analysis of the entire data set is desired? Without sampling?<\/em><\/p>\n

The solution to the puzzle: Custom Tables<\/h2>\n

This is exactly why Google offers its Analytics 360 customers the new 360 feature called Custom Table.<\/p>\n

With custom tables orcustom tables<\/strong>can create custom reports with a combination ofDimensions, metrics, segments<\/em>andFilter<\/em>created by Google Analyticsdaily without sampling<\/strong>processed. This requires a special configuration in the Google Analytics Admin Interface.Important:<\/strong><\/u>A custom table deliversno new report<\/u>in the Reports interface. A custom table is a configuration whereby all data that matches that specific configuration uses the table by default.<\/p>\n

How To: Create a custom table<\/h2>\n

Custom tables are onProperty level<\/strong>created and configured. You can do this in the administration interface underProperty -> Custom tables<\/em>, after selecting the desired property,maximum 100 custom tables<\/strong>create. Once custom tables have been created, they cannot be subsequently edited. In order to make changes, it must be deleted and recreated.\"\"Theconfiguration<\/strong>works the same as when creating onecustom reports<\/em>: You choose onenames,<\/strong>the one you wantView<\/strong>and a short oneDescription,<\/strong>selects up to 6Dimensions<\/strong>and up to 25Metrics<\/strong>from, addsSegments<\/strong>andfilter<\/strong>added and saves it.\"\"The data is then pre-calculated in the Google Analytics backend and made available after 48 hours at the latest. About the administration interface and the selection“Show custom table”<\/em>The unsampled report can be accessed in the action drop-down. It is also possible to delete the user-defined table here.\"\"Important:<\/strong><\/u>Custom table queries cannot only<\/u>can be created via the Google Analytics Admin interface, but also via theGoogle Analytics API e.g. via Google Spreadsheets.<\/strong>The Custom Table queries are based on normal GA Reporting API queries – namely on exactly the same settings (dimensions, metrics, filters, segments) that are also queried in the Custom Table. Google is taking actionin the background automatically<\/strong>access the custom table if it is available. If no custom table is available, Google automatically grabs the sampled data from GA.<\/p>\n

Custom table analysis<\/h2>\n

The analysis described in the introduction looks like this using Custom Tables:\"\"Very elegant! Very clear!And above all: Unsampled!<\/em>\"\"<\/p>\n

Summary and overview<\/h2>\n

Custom tables are used to obtain unsampled reports for large scale analyses.<\/p>\n