If you have been following the mobile app ecosystem closely you know that there are more and more apps that are relaying on advertising revenues to create profitable businesses. Zynga reported significant growth in ad revenue in their recent financial statements and leading speakers in the industry have talked about the trend towards ad based monetization in industry events.
This situation leads to new challenges in ROI measurement. Apps need their LTV to exceed the CPI and LTV calculation require you to know the ARPDAU at some point. Lets see how ARPDAU is calculated in different monetization methods.
Measuring ARPDAU for Apps that use IAP
If your app only uses IAP for monetization, your life is quite easy. To get the ARPDAU, simply divide the daily revenue by the number of DAU. both parameters can be obtained from your analytics platform or in-house BI. Figuring the ARPDAU in a specific segment, cohort or traffic source is easy since the data is available for each user so all you have to do is repeat the exercise for the group of users that are in the segment.
Ad Revenue per user is not available for ad-supported apps
The life of ad-supported apps is more complex. Here are some of the challenges:
- Using multiple ad-networks means that the revenue information needs to be collected from multiple dashboards
- The data you can get from the ad-networks is aggregated to the country/day level – no data per user is available
- Understanding who are the users who click on the ads is very hard and getting install data is almost impossible
- The 90% of the users who don’t click don’t generate any revenue on CPI or CPC campaigns (most of the campaigns today)
Using an average leads to errors
Mobile app companeis have been using an average to calculate the ad revenue per user, LTV, and ROI. This method is pretty simple – they take the revenue generated in a specific country in a given day and divide by the amount of users that day to receive the ad revenue per user. This of course, assumes that all users are contributing the same amount of revenue which is very far from the truth. In reality, only about 2% of the users in a given app actually go and click on the ads and install the advertised apps. This means using an average is wrong 98% of the time.
Counting impressions per user is also not accurate
More advanced app publishers have implemented ways to count the number of impressions served per user and per segment and have been using that to calculate LTV. This method is also wrong. In most ad-networks the revenue is driven by the CPI and CPC campaigns and therefore the impressions are not a good indication of revenue. For example, a user with 100 impressions could have generated $0 while a user with 2 impressions who also clicked and installed could have generated $2.
Be sure to check our latest Ad LTV Benchmarks report.
If you want to get your ad revenue right 100% of the time you can now. Check out SOOMLA – Ad LTV as a Service.