Mobile Phone Application Data ("App Data") correlates to certain company financials and given its high (daily) measurement frequency, it can be used to regularly update and predict estimates for these metrics.
The following article, using Isracard Ltd. as an example, explains how:
to identify relevant Apps
to evaluate the quality of the App Data
to predict financial metrics as stated in company quarterly reports using App Data
38% of Isracard Ltd.’s revenues come from credit card holders
Isracard Ltd. is a credit card company. It is active in the fields of issuing and clearing various branded debit cards and in financing. The company charges users of its services fixed annual fees and commissions on transactions.
In 2018, 44% of Isracard Ltd.’s revenues came from businesses while 38% of its revenues came from credit card holders, this is important to note in the context of App Data.
Isracard’s main Android apps
Isracard's Clients can use mobile apps published by Isracard in several ways. The table below shows the main Isracard Android apps (this info can be easily found on the Google Play Store or the Apple App Store):
Isracard Ltd. has two apps with over 100’000 downloads for Android: “Isracard” and “American Express”. In this analysis we only use data from the “Isracard” app. Note that this app is intended for Isracard credit card holders. The analysis can definitely be improved by including the “American Express” app as it also has a significant number of downloads, and Isracard's Apple apps. Its unclear if the data for the other Android apps will be of good quality given their low ratings and downloads numbers.
Evaluating App Data Quality for the “Isracard” app
Apps need to be ranked in the public Google Play Store Ranking for app data providers to be able to estimate downloads. Thus we need to know how often the app is ranked in the period we are interested in.
To understand this we first purchase historical app rankings data for the Android Isracard App in the Israeli app store (as most of Isracard's clients are based in Israel):
In the period of analysis the app was ranked an average of 97% of days each quarter (in the Google Play Store “finance” category ranking) indicating that download data will be available for almost all days. We thus move on and purchase download data for the Isracard App in Israel.
The table below shows the app download data combined with Isracard's financial metrics.
The estimated precision data (supplied by this data provider together with the downloads data) gives us a further indication of thee quality of the download estimates. In this case the average precision is 70% per quarter (note that the precision becomes higher in later quarters). The data thus looks to be quite reliable.
The plot below shows how the daily app download data and precision evolve over time for the Isracard App.
App Downloads predictive of both Revenue from Card Holders and Total Revenue
Finally we test the predictive power of the Change in App Downloads against the Change in Revenue from Credit Card Holders (as the app is in fact intended for Card Holders) and the Change in Total Revenue.
The image below gives the results of these two regressions:
The results show that App Downloads are a significant predictor of both Isracard Ltd.'s Change in Revenue from Credit Card Holders and Isracard Ltd.'s Total Revenue:
P-values are shown to be 0.48% and 3.5% respectively,
correlations are 73% and 59% respectively
R squared values are 53% and 34% respectively
As expected Changes in App Downloads are much more informative of Change in Revenue from Card Holders, but also informative of Change in Total Revenue.
Further analysis using business intended and iOS apps should be performed to complete the picture.