Client and goals
Web promo and Ukrsibbank Paribas Group started working in September 2019 with the B2B direction.
In 2020, Ukrsibbank released an updated application where it introduced new functionality that allowed users to become bank clients and make all financial transactions on their smartphones.
In May, Web promo specialists were tasked with developing and implementing a strategy to attract new customers through the UKRSIB Online mobile application.
The main KPI was the number of attracted customers, as well as installs, which served as the first user interaction in the card design chain.
Advertising of the mobile application “UKRSIB Online” on the Internet
Initially, we started working with standard sources for app promotion:
- Google Ads – ads are displayed on the Display Network, Search Network, YouTube, Discover, Google Play;
- Facebook Ads – placements have been adjusted and we settled on Facebook/Instagram/Messenger.
In the process of work, we also carried out adjustments for placements. The most effective was the newsfeed and stories on Facebook/Instagram.
After launching the main tools, we set up an additional source of installs – Apple Search Ads. This tool allows you to show ads to users in the App Store.
Google Universal campaign
The Google Ads system allocates a separate type of Universal Campaign for apps. This tool is automated. At first glance, it seems that there is no way to optimize advertising campaigns and you need to put up with the results that you get.
In reality, there are always optimization options, and app campaigns are no exception.
In the task from Ukrsibbank, certain geography of displaying advertising campaigns was highlighted. Based on this, we decided to segment campaigns by geography, since each region has different competition:
- Kyiv;
- Cities with a population of over 700,000;
- Cities with a population of over 100,000;
- Cities with a population of less than 100,000;
This approach allowed us to obtain data for each GEO and quickly redistribute budgets among the most effective ones, which allowed us to achieve good results.
The most expensive installations turned out in Kyiv. This is due to the high competition within the Google auction between advertisers since this region has the most solvent audience.
In comparison with Kyiv, we obtained the following data:
- In cities with a population of more than 700,00, CPI is cheaper by -11.19%
- In cities with a population of more than 100,00, CPI is cheaper by -13.58%
- In cities with a population of less than 100,000, CPI is cheaper by -26.76%
With app campaigns, you can use up to 5 titles, and up to 5 descriptions, and we recommend uploading up to 20 images, up to 20 videos, and up to 20 HTML5.
Ads are generated automatically based on the objects and text options you upload, as well as your content on Google Play and the Apple App Store.
In our work, we decided to additionally segment campaigns into ad formats – images and HTML5 banners.
The cost per install for image campaigns is -10.99% lower than for HTML5 campaigns. This data allowed us to effectively optimize advertising campaigns. As a result, it was decided to reallocate investments in favor of the campaign with images.
At this stage, the client is preparing videos that we will test according to the same principle and then make decisions about the effectiveness of certain formats.
Facebook Install Apps
The campaign structure in Facebook Ads was built according to the audience principle:
- Interests – portraits of the target audience were worked out;
- Look-alike;
- Client lists;
- Remarketing audiences;
- Wide audience.
This structure allows us to effectively allocate budgets between audiences and achieve better results.
One of the key features of Facebook promotion is tracking how often ads are shown. As we know, Facebook pays per impression (even if you choose to pay per click, the system still calculates the pay per impression, this is the peculiarity of Facebook Ads). If you do not work with the frequency metric, then your CPM will grow. This will lead to an increase in CPI, which directly affects the number of installs.
The solution with which we close this complexity is to work with an acceptable frequency of up to 2.5, after which the advertising campaigns are completely restarted with new banners and texts.
In their work, Web promo specialists regularly analyze creatives, which allows them to provide the client with the most successful solutions and give recommendations for creating similar banners, as well as writing texts with an emphasis on one or another USP.
We constantly analyze advertising campaigns. When there is a continuous upward trend in frequency and CPI, which leads to a decrease in the number of installations, we ask the client for new banners. We can highlight those that were the most successful and brought the best results in the promotion of the UKRSIB Online mobile application:
This approach to working with Facebook Ads allows you to:
- Quickly identify the most effective audiences and invest more in them. As a result, it is possible to reduce CPI and get more installs within the budget;
- Track frequency is one of the key metrics to achieving the best result. Conduct deep analysis of creatives and launch new ones based on them, thus reaching the target audience with a high-quality and attractive offer.
Apple Search Ads
Apple Search Ads (ASA) is a tool for advertising an app in App Store search. Ads appear at the top of search results.
According to the App Store:
- 70% of App Store visitors use search to find apps;
- 65% of downloads happen right after a search in the App Store;
- 50% is the average conversion rate for Apple Search Ads.
ASA features:
- Quick launch for the IOS platform, since there is no need to implement the SDK as it is done in other sources;
- Unlike Google and Facebook, a user enters the App Store with a clear intent to use the app.
We have worked out the structure of advertising campaigns, which allows you to receive large impressions for targeted queries using the “search match” function. This tool helps you automatically show your ads to users who are looking for apps similar to yours.
After receiving the statistics, we started optimizing advertising campaigns and made the following decisions:
- Disable campaigns without the “search match” function – they received little traffic and installations were more expensive;
- Suspend ineffective keywords and advertising campaigns, as well as reallocate the budget to more effective areas;
- Expand flagship advertising campaigns by collecting new keywords;
- Adjust your bids at the keyword level to the optimal ones to get the most installs at an affordable cost.
Comparison and evaluation of the effectiveness of promotion channels
In our case, we considered working with three traffic sources for the UKRSIB Online application.
The largest number of installs was received from the Android platform – 85.7%.
After receiving statistically significant results, we decided to scale up advertising campaigns.
After analyzing the activities, most of the budget was invested in Android, as the cost per install was -71.01% lower than in IOS.
This trend continued with the scaling of advertising campaigns.
Comparing promotion channels, it is logical to do this in terms of platforms:
Android. is the lowest cost per install in Google Ads -87.2% compared to Facebook Ads.
iOS. is the lowest cost per install in Apple Search Ads at -75.4% compared to Google Ads and -75.81% compared to Facebook Ads.
As a rule, installs for the IOS platform are more expensive from standard Google Ads/Facebook Ads sources. But in working with Apple Search Ads, this disadvantage is compensated and it is possible to get installations much cheaper than from other sources.
The key factor in working with the application is determining the qualification of installs from a particular source/platform and optimizing advertising campaigns based on an analysis of the quality of the received installs.
This analysis can be carried out using analytical systems for applications. We recommend using “Appsflyer” as a system that allows you to:
- Conduct multi-channel analysis;
- Display data from multiple channels in one interface;
- Analyze audience behavior;
- Provide Real-time reporting;
- Integrate it with many other services;
- Use deep-linking;
You can learn more about mobile app promotion in our case studies:
