Date live in Priori Data: May 08th, 2018
Model(s) updated: Downloads, DAUs, MAUs, ARPDAU
Platform(s) affected: Google Play
What did the update do?
Similar to the model update for iOS in April 2018, for this release we re-fit the download model for Google Play on historical data provided from additional publisher partners.
In addition, we have improved our modeling methodology by fitting a more precise and stable formula depending on data availability, as well as introducing a new approach to handling day-of-week effects. These changes allow us to more accurately capture the downloads curve of smaller markets.
Finally, we have made adjustments to our post-processing approach to better smooth daily estimates. This helps remove erratic or inconsistent behavior in daily trend lines for apps.
As a result, we have made significant improvements in the MAPE (Median Absolute Percentage Error) for all countries, as well as for almost all categories. The following tables show the download accuracy improvement in MAPE for the top 10 countries.
Top 10 Countries in terms of Download Accuracy Improvement:
% Improvement in Median Absolute Percentage Error, November 2017 - January 2018
Impact on usage estimates
Since our DAU, MAU and ARPDAU estimates are strongly dependent on the downloads predictions, we expect this update to have a positive impact on our DAU, MAU and ARPDAU estimates too.
Overall, better models means better accuracy. It's important to note that the changes in our estimates might affect some apps, publishers, and categories more dramatically than others. If you have questions about how a specific subset of our platform has been impacted, please contact us (firstname.lastname@example.org).