Search
Clear search
Close search
Google apps
Main menu

Troubleshoot app statistics problems

If you're having trouble with your app statistics data, the information below may help.

Why did my "Daily installs" go down?

Here are some app changes that might cause a decrease in installs:

  • New permissions that require users to manually accept changes
  • Changes to country or language distribution
  • Changes in advertising
  • Changes in app placement within the store
  • Carriers pre-installing your app (system apps are only counted in statistics after an initial update through Google Play)

If you believe none of these scenarios apply to you, contact our support team.

There's a problem with the statistics report I downloaded.

If you're unable to open the report you downloaded from the Play Console, contact our support team.

View updates to Google Play statistics

December 5, 2016: Increased refresh rate for "daily installs & uninstalls by user" data

Starting on December 5, 2016, we'll reduce the latency associated with updating your app's "daily installs by user" and "daily uninstalls by user" data on the Statistics page. This means you'll be able to access this data sooner than you previously could.

For some apps, you may notice a slight shift in "daily installs by user" and "daily uninstalls by user" data on this date.

November 30, 2016: Change to store listing experiment results and your app's current installs data

Store listing experiment results

Starting on November 30, 2016, results for any new or pending store listing experiments will be based on the metric, "Installs on active devices."

If you started an experiment before September 1 that's scheduled to end after December 1, we'll provide you with data going back to September 1. For the most accurate results, we recommend ending any experiments that began before September 1 and creating a new experiment.

Current installs data

Your "Current installs by device" and "Current installs by user" data will be available until November 30, 2016 on the Statistics page and in exports. For data after November 30, use "Installs on active devices" instead.

For more information, see the "introduction of installs on active devices" section below.

September 29, 2016: Introduction of "Installs on active devices"

We're introducing a new metric called "Installs on active devices." Initially, you'll see this metric on your app's Statistics page, with data going back to September 1.

Later this year, the new metric will replace "Current installs by device" and "Current installs by user" in all current reports. You'll still be able to access historical data for the replaced metrics in charts and CSV downloads.

Why are install metrics changing?

"Installs on active devices" shows how many devices that have been online at least once in the past 30 days have your app installed.

Your app's "Current installs by device" and "Current installs by user" data has included devices that were last active up until several months ago. By switching to "Installs on active devices," we're able to focus on devices used within the last 30 days to show you recent user engagement with your app.

Impact on your metrics

Since "Current installs by device" and "Current installs by user" included a longer activity window, it's expected that you'll see lower numbers for your app's "Installs on active devices" data.

Timeline

We'll keep you updated on the timeline for this change within the announcement section of your Play Console.

  1. Go to your Play Console.
  2. In the top right corner of any page, select the bell icon Announcement.

Here's what you can expect:

Timeframe Impact on data
9/29
  • You can view "Installs on active devices" as an additional data point on your app's Statistics page.
11/3
  • "Installs on active devices" will become the default metric shown on your app's Statistics and All Applications pages and the Play Console app.
  • "Current installs by device" and "Current installs by user" will remain available on your app's Statistics page.
11/30
  • "Current installs by device" and "Current installs by user" will stop updating.
  • Historical data will remain available in charts and CSV exports.
November 29, 2015: Change in "Current installs by user," "Current installs by device," & "Total installs by user"

Starting on November 29, 2015, we will introduce new criteria for counting user and install metrics to account for users and devices going inactive. This change will affect the following metrics: "Current installs by user," "Current installs by device," and "Total installs by user."

As we make these changes, you may notice that these metrics decrease over time for some apps.

June 27, 2015: Update to Google Play installation statistics

On June 27, 2015, we updated the install statistics on your Play Console to use more accurate and reliable data.

As a result, some developers may see small changes in their apps’ install and uninstall data on their apps’ Statistics pages.

May 28, 2014: Counting of preinstalled apps and multiple users on the same device
On May 28, 2014, we updated our statistics to make corrections to how we counted factory resets for preinstalled apps and apps that are used by multiple users on the same device. Most developers will see minimal impact to metrics, though certain developers will see significant impact to Daily Device Installs, Daily Device Uninstalls, and Current Device Installs.
August 5, 2013: Update to better identify active devices
As of August 5, 2013, we are updating our stats algorithm to better identify active devices. This update will improve the accuracy of the installation metrics for active device installs and active user installs. When the new algorithm is rolled out, you may notice changes in some of the metrics associated with your app. Many apps will be affected by this change, however we expect a difference of less than 2% for most apps.
December 2012 - January 2013: Possible decline in total user installs and total device installs for paid apps due to data migration
In December 2012 - January 2013 developers may notice a decrease in the total user install and total device install counts of the paid apps displayed in their Play Console. This is the result of an ongoing data migration, and is the true count for your app. The migration will be done in stages and will continue through January 2013. This migration will not affect previous developer payouts.
April 4, 2012: Better identification of Android platform version
As of April 4, 2012, we have improved the accuracy of statistics regarding the Android platform version number. As a result, you will see fewer devices classified as “Other” or “0” and more devices classified accurately.
February 13, 2012: Inactive Android devices
As of February 13, 2012, we have removed data about Android devices that are no longer active, which corrected active users and devices for many apps.
January 28, 2012: Total user installs undercounting
As of January 28, 2012, we have resolved an issue that caused undercounting of total user installs.
May 2, 2011: Change in the number of total installs
As of May 2, 2011, we've changed the basis for calculation of install numbers provided in Play Console. Previously, app installs were calculated per user and did not account for certain uninstall activities (like when a device was factory reset). Installs are now counted per device, with duplicates more effectively filtered out. This offers you a more accurate measure of your app distribution in the Android ecosystem.

You may notice that this change has resulted in a one-time decrease in the number of total installs for your apps. We apologize for any confusion this may have caused.

October 18, 2011: Decrease in the number of active installs
As of October 18, 2011, developers may notice a decrease in the active install counts displayed in their Play Console. Some app updates were being counted as device installs. We are now using a different calculation methodology which does not incorporate installing an app update. Going forward the active installs metric will correctly reflect the device installations. Please note that total install counts were not affected.

Was this article helpful?
How can we improve it?