App Monetization, Tips and Advice

Gazeus – Measuring Ads Allowed Us to Save 66% of Our Budget

How Gazeus saved 66% of their budget by measuring ads

At SOOMLA we are always advocating being more data driven when it comes to mobile app monetization. This way we are happy whenever we see a publisher stepping up their data game and starting to measure their monetization. In an earlier post we showed how Kongregate doubled their traffic by measuring ads and in this post we will see how Gazeus saved 66% of their budget.

About Gazeus

If you don’t know Gazeus I highly recommend watching the video below. The presentation is given by Paula Neves, CMO @ Gazeus.

 

 

Here are a few key points to know about Gazeus:

  • Owner of the Jogatina brand that is known from web/PC games
  • 40+ Titles
  • 9M MAU in mobile
  • 99% of mobile revenue coming from Ads
  • Strongest Geos – US, Canada

Measuring Ads – Critical and Challenging Problem

Since 99% of the revenue is coming from Ads, says Paula: “LTVs are lower and UA has to be spot on”. She then explains that they had ROI calculations only at the app level as a whole but not per campaign. On the product side, they didn’t have sufficient feedback loop for product to know if the features they are making are good. For these reasons, ad measurement is critical for Gazeus. In the video, Paula explains that her background is in the eCommerce domain where she took LTV for granted but when she came to Gazeus, she discovered that an entire infrastructure has to be created just to receive an answer for this question. This root cause of that is, of course, lack of user level data from the ad-networks.

CASE STUDY ON ADVERTISERS CHURN & eCPM

The outcome – 66% of marketing budget saved without sacrificing growth

Skipping to the bottom line quickly. Paula explained that the ability to measure ads allowed them to be much more focused in their UA efforts. In one example she gave, having the granular ad revenue per user data allowed them to chart the LTV curve for each cohort. By comparing the curves of iPad vs. iPhone users coming from different campaigns they were able to save a big portion of their ineffective marketing spending. The bottom line is that they were able to spend as little as 1/3 of what they were spending and still get the same revenue growth.

Gazeus ad measurement infrastructure and SOOMLA

In the middle part of the video, Paula explains how they built their ad measurement infrastructure. They had to collect API level data from each one of the ad-networks. For each one, different types of fields and dimensions are provided so normalizing that and the timezones is another challenge altogether. From the client side, they had to set up an event to be fired every time an ad is served. The server than allocates the average eCPM of the previous day for the relevant country and ad-format for each impression and multiplies by the number of impressions to generate user level revenue data.

Of course, a publisher doesn’t have to go through all this in order to get user level ad revenue. They can simply work with SOOMLA to get the data through API, directly into Amazon or via SOOMLA Dashboard. For Gazeus, they decided to go with their own solution since they use a few unique networks in Brazil that SOOMLA doesn’t support as Paula explains but for most publishers it’s not an issue. Here is a table to help you evaluate both options.

Build In-house Using SOOMLA
Initial Effort Up to 6 man months 1 week
Calculation Method Average eCPM Method* True eCPM Method**
Ad Network Support Tailored The 25 Most Popular Networks
Maintenance Internal resources Outsource to SOOMLA
Extra features N/A Advertiser identification, ad whale lookalikes, postbacks
Cost (for 5M MAU) Increase in hosting bill can reach $2,000 $2,500 to $3,500

*Average eCPM Method

  • In this method, the ad LTV system queries the API of each ad-network to retrieve the average eCPM of the previous day for each country and ad-format.
  • At the same time, the client side of the app reports to the ad LTV system each time an impression is shown to a user.
  • The ad LTV system calculates the revenue for the user by multiplying the number of impressions the user made in each ad-format by the average eCPM of that ad-format/country in that day and than dividing by 1,000. Than it sums the results for each ad-format for that user.
  • The process is repeated each day and data is aggregated on a user level basis and on a cohort basis.

Putting it in a formula – this would be the calculation of the revenue for user x (Imps is short for impressions here):

**True eCPM Method

  • In this method, the ad LTV system queries the API of each ad-network to retrieve the average eCPM of the previous day for each country and ad-format in addition to other dimensions that are available.
  • At the same time, the client side of the app reports to the ad LTV system multiple events for each impression: impression event, clicks, installs, ad interactions, advertiser identity, business model, bid levels.
  • The ad LTV system leverages advanced algorithms to deduct the true eCPM of each impression based on the data collected from the client side as well as the averages reported through the APIs. It then calculates the revenue for each user by summing up the contributions of each impression he viewed.
  • The process is repeated each day and data is aggregated on a user level basis and on a cohort basis.

The formula below shows the revenue calculation for user x where ‘n’ is the number of impressions he made in that day.

Feel free to share:
Share this Story
Load More Related Articles
Load More By Yaniv Nizan
Load More In App Monetization

Leave a Reply

Your email address will not be published. Required fields are marked *

DOWNLOAD OUR RECENT REPORT

Join 8,218 other smart people who get email updates for free!

We don't spam!

Unsubscribe any time

Categories

LAST STEP - ENTER EMAIL TO GET THE REPORT



By proceeding, you agree with our terms and conditions and privacy policy.

LAST STEP - ENTER EMAIL TO GET THE REPORT



By proceeding, you agree with our terms and conditions and privacy policy.

LAST STEP - ENTER EMAIL TO GET THE REPORT



By proceeding, you agree with our terms and conditions and privacy policy.

LAST STEP - ENTER EMAIL TO GET THE REPORT



By proceeding, you agree with our terms and conditions and privacy policy.

LAST STEP - ENTER EMAIL TO GET THE REPORT



By proceeding, you agree with our terms and conditions and privacy policy.

LAST STEP - ENTER EMAIL TO GET THE REPORT



By proceeding, you agree with our terms and conditions and privacy policy.

LAST STEP - ENTER EMAIL TO GET THE REPORT



By proceeding, you agree with our terms and conditions and privacy policy.

LAST STEP - ENTER EMAIL TO GET THE REPORT



By proceeding, you agree with our terms and conditions and privacy policy.

LAST STEP - ENTER EMAIL TO GET THE REPORT



By proceeding, you agree with our terms and conditions and privacy policy.

Join 8,218 other smart people who get email updates for free!

We don't spam!

Unsubscribe any time