Discrepancies in Google Analytics Results

While most of the time, Supermetrics matches exactly the figures that the Google Analytics web interface displays, sometimes there are discrepancies. This guide will cover the most common reasons: data sampling and UI latency.

Quick links to sections below:

Reason #1 - Data Sampling

Google may sample the results it delivers to Supermetrics, which can cause differences between the values displayed by Supermetrics and Google Analytics. The magnitude of the discrepancies is usually in the range of 1-4 %. Sampling occurs most often in queries that include an advanced segment, very large data sets, large numbers of session data, or where the data is split by multiple dimensions. 

For more information about data sampling, see the Google Analytics Reporting API documentation and the General GA information article.

How to Check for Sampling

To see whether Google has sampled your query results, check the appropriate section based on the Supermetrics product being used:

Google Sheets: Open the SupermetricsQueries tab and see whether the "Results contain sampled data" row has the value TRUE (sampled) or FALSE (not sampled)
Example SupermetricsQueries sheet showing a sampled query marked as "TRUE"

Excel: Open the SupermetricsQueries tab and see whether the "Results contain sampled data" row has the value TRUE or FALSE

Data Studio: Sadly, there is no way to tell in this product. Jump forward to the directions on how to avoid sampling to see if that helps

Data Grabber: The report status will note it was sampled in the output (if this is not listed, it is not sampled).
Orange box showing sampling text for Data Grabber.

For Google Sheets and Excel, there is also an option to "Add note to query results showing whether Google has used sampling" which will display it at the top of the results.
Example unsampled data in Excel using the option to add a note to the results.

How to Avoid Sampling

In most cases you can avoid sampling by enabling the option to "Try to avoid Google's data sampling" in the product.

Google Sheets: The option is under the Options section of the sidebarAvoid sampling option in Google Sheets

Excel: The option is selectable under the Options section of the sidebar
Avoid sampling option in Excel

Data Studio: The option is in the general configuration section of the data source (click icon to edit data source, then EDIT CONNECTION to get back to this area)
Avoid sampling option in Data Studio

Data Grabber: The option is listed on the Analytics Tab with the other query settings.

Be aware that this can slow down your query and increases the request amount needed for the query, leading to potential quota issues.

Queries that are large/complex enough will still be sampled by Google, even with this setting enabled. If you continue to get data sampling even with the "Avoid Sampling" option enabled, try some of these suggestions to simplify the query:

  • Reduce the number of dimensions used
  • Reduce the date range of the query
  • Move any filtering into the Google Analytics view instead of using the filter options in the query
  • Take advantage of the Combine New Results with Old setting (Google Sheets only)
  • Use fewer views at once
  • Use fewer segments at once

There is also an advanced setting SAMPLING_RATE that can be used with products that support those, to show what % of the data was sampled, which may give a better understanding of Google is doing when it does sample.

Reason #2 - API and Web UI Out-of-Sync

The Google Analytics API and the Google Analytics web interface are not always totally in sync. Supermetrics products rely on the Google Analytics API, which may in some cases deliver figures which differ from the Analytics web interface. Small differences can be seen every now and then, but sometimes bugs in Google's end cause major discrepancies. 

If you suspect something like this may be affecting your reports, there are a few places where you can investigate the issue:

  • Analytics status dashboard. Here you can see the status of the Analytics API and other Analytics services, bugs may be shown here.
  • Analytics API notify list. Serious issues may be posted here by Google.
  • Analytics Query Explorer. This is Google's own API tool, which you can use to confirm the results showed by Supermetrics. If this tool agrees with Supermetrics and differs from the Analytics web interface, then it indicates a bug in Google's end you should report to them. On the other hand, if Google's API tool and the Google Analytics web interface match, and Supermetrics shows something different, that indicates a problem in our end; in this case, please contact us immediately.

Engaging Support About Discrepancies

When you see a discrepancy, before notifying engaging Support about the issue, please follow these steps FIRST:

  1. Verify the issue is not caused by data sampling, as described under Reason #1 above. If it is caused by data sampling, follow the directions in that section for how to try to avoid it. We cannot do anything about data sampling otherwise.
  2. Try the exact same query in the Analytics Query Explorer tool as noted in Reason #2 above. If the results returned by the Query Explorer match Supermetrics, the issue is with the Analytics UI and you need to contact Google for further support.

If you if have ruled out the above reasons for the discrepancy and believe it to be an issue with Supermetrics, then submit a support ticket with the following details:

  • Confirm that you have followed the steps in this article completely.
  • Include screenshots showing the differences between Supermetrics results and Analytics. Make sure the date ranges are exactly the same in each screenshot.
  • Include the link to the report with the discrepancy issue, make sure support@supermetrics.com has EDIT access to the file.
  • Include a copy or screenshot of the results from the Query Explorer tool showing the API returns different results.

Did you find it helpful? Yes No

Send feedback
Sorry we couldn't be helpful. Help us improve this article with your feedback.