Background

The client contacted us with the need to advertise the newly created Bond.ua application, a tool for buying domestic government bonds (better known as “military bonds”).

Our Goal

For many Ukrainians, the opportunity to buy domestic government bonds was news. But such a new product received significant media support in the market. ThThe media and opinion leaders presented the bonds as a profitable financial instrument, and their purchase as a patriotic act. So, despite the novelty of the product, the client did not need to warm up the market – the state had already told the public about the product.

We, as marketers, created the first advertising campaigns for the application, and our goal was to gain the initial audience of the app. And the main KPI of campaigns was the number of installations.

Our Approach

There can be no detailed targeting in app installation campaigns, neither by age, nor by gender, nor by audience. So the main target for us was geo. The application was promoted in Kyiv, Odesa, and Dnipro. Also in Lviv, Chernivtsi, Ivano-Frankivsk, Zakarpattia, Khmelnytsky and Ternopil regions. The regions were suggested to the client based on the situation in the country- we focused on the conditionally safe regions in the spring of 2022.

Tools

The tracking was done through Firebase, which was installed on the client’s side and by his own efforts. On iOS, the conversion was counted only after the user opened the application. That is, it was not about the installations themselves, but about the deeper convention of the First Open. Because of this, we did not see the actual number of installs in the advertising campaign, but rather the number of users who installed the app and opened it. At the same time, there was no such confusion on Android, and we saw information about installs.

Earlier, we already had problems with Firebase tracking when the system claimed that there were no iOs users in Cyprus at all. Then we were able to find “lost” users after changing the tracking system. Read more in the case study: Firebase VS Adjust: How Display Campaigns Can Suffer Because of Data.

Google Ads become the main tool for creating promotion. We decided to choose it as a guarantee of a proven result. The client had not launched yet, and the product was new, so we decided not to experiment, but only to test the most likely hypotheses. In addition, Google Ads campaigns for installs are adaptive. You can set options for titles, descriptions, and creatives, and the system itself, observing users, will select the most effective combinations of options.

We also planned to launch advertising on Facebook, but we had to postpone this idea. Meta does not allow promotion of applications that are in beta, as it is not a fully finished product for end-user promotion and can damage reputation. Also, a mandatory requirement for promotion is to set up the Facebook SDK- a tool that is used to set up an advertising campaign and obtain an Application_id. Together with the client, we agreed on the nuances for the advertising campaign on Facebook and will share this experience later.

By the way, we have the manual on how to share access to a Meta ad account.

Results

In total, during the three weeks of the campaign ( from mid-April to early May), we reached almost 2 million users, of whom more than 8,000 installed and opened the app. The cost per install throughout the campaign was lower than planned, while the number of installs exceeded expectations by half.

In terms of conversion rate, Lviv region and Kyiv drove most of the traffic to Android. They accounted for the largest and approximately equal share of impressions and installs. At the same time, the CPA for all cities and regions was approximately the same, differing from each other by 1 cent.

On iOS, we saw a similar picture – most conversions came from Kyiv and Lviv region. But the cost of conversions from region to region differed by almost half. The average cost per conversion on iOS across all geos was ten times higher than the cost per conversion on Android.

In general, the number of installs increased and their cost decreased throughout the campaign. This is exactly what should happen when your ads are set up by professionals.

Conclusions and recommendations

1. According to the results of the advertising campaign, we exceeded our targets for almost all the metrics we set for ourselves: impressions, clicks, CPC (iOS) and CPA (Android), although the CPA for iOS was 12.64% higher than planned. We exceeded the main KPI, the number of installs, by 72.24%.
2. In future advertising campaigns, we recommend expanding geo-targeting to reach a larger audience. Regional centers such as Uzhhorod, Lutsk, Lviv, Ivano-Frankivsk, etc. should be singled out. Expanding the audience will provide a larger statistical sample to draw further conclusions based on the data. But, of course, we need to take into account the situation in the country and not run ads in areas and cities close to hostilities.
3. We recommend testing Apple Search Ads in future campaigns to increase iOS performance. This should allow you to influence the price, reach, and conversions among users of this platform.
4. It is also worth launching campaigns optimized for deeper conversion – in-app purchases of bonds.
5. Before you can promote an app on Facebook, you need to configure and link apps to Facebook, configure the SDK, and bind application_id. If the application involves monetization, the sale of anything in the app, you need to properly configure these internal payments. Without fulfilling all the requirements, it will be impossible to create an app advertising campaign on Facebook.

If you’re looking for an analytics solution for your mobile apps, you’ll be interested in our Firebase & Adjust comparison case study.