The vast majority of your players will be receptive to some kind of engagement. The game developer’s job is to pinpoint and provide what works, for which players, through sophisticated messaging and A/B Tests.
Casualino used our powerful Engage toolset to run experiments on and optimize a number of in-game store offers, resulting in revenue uplift of up to 200%.
We had lots of examples to choose from but, for the purpose of a fair case study, we are looking at a small selection of campaigns comprising especially high numbers of participants.
Belote is a 32-card game, popular in many countries across Europe. The app version Belot.bg was developed by Casualino for the Bulgarian market and is available on iOS, Android, Windows mobile and Facebook.
During the 3 days preceding New Year 2019, a variety of campaigns were served to players upon opening the game. The control group received no offers at all, and our 2 target groups received different experimental variants of offers from the in-game store.
For our target groups, at first open, a 99.99 BGN (Bulgarian Lev) offer was served to players across the board.
If and when the initial offer was not taken up, another deal was promoted an hour later – either in the form of a free ‘gift’ or a discount ‘sale’ offer.
To work out the impact of these variants, we compared the Average Revenue Per Daily Active User (ARPDAU) for the period, with the ARPDAU of the Control Group used as base of 100. As you can see in the chart below, the recipients of the gift and the sale promotions generated greater revenues than the control group, by an average of 46%. It is interesting to note that the two variant messages, while different, delivered similar returns.
Iterate to optimize
Later in the year, at the start of September 2019, a bonus promotion was released in collaboration with an online trading service for gifts and services. Female and male players would see a different image for the campaign upon starting Belot.
In Image Set A, we see that neither variant outperformed the control group in this instance.
However, later the same month, a similar campaign was launched using alternative collateral for both “him” and “her”.
When exposed to Image Set B, the average revenues for target groups were significantly greater than that of the control group – by a whopping 200%. This uplift demonstrates the essential value of iterative experimentation in games – one attempt is never enough.
Interestingly, Image Set A was also being used to engage other player segments. One such group comprised players who had viewed someone else’s profile within the game.
In this instance, the images did appear to improve outcomes when compared to the control group – by an average of 90% across both female and male versions.
Again, these results highlight the difference between player groups and the need of bespoke targeting. An offer that doesn’t perform well in one instance is not necessarily a bad offer – it might do wonders if presented to a different, specific, audience.
The bottom line
With their evergreen curiosity and desire to optimize, Casualino uses our flexible engagement tools exactly as we built them to be used. We spend a great deal of time talking up the importance of targeted engagement, but that’s only half of the equation. A continuously experimental approach, creating and testing hypotheses, is the only way to guarantee optimization and do your game justice.
By being disciplined with their control groups, and meticulously testing the possible variants of their offers, they avoided the pitfalls that typically prevent games from reaching potential. In this case, the uplift speaks for itself.
“We often see personal biases and pride get in the way of building better games, but not here. By adhering to best practices in messaging and optimization, the team at Casualino executed textbook A/B Tests that ultimately transformed the impact of their offers. We’re excited to see what they do next!”
Russell Young, Analytics Consultant, deltaDNA
“We’re extremely happy with the doors that working with deltaDNA has opened up in the last 3 years. The extensive engagement tools allow us to experiment with segment targeting and help us deliver the best possible experience to our users. The A/B testing tools are an integral part of our engagement strategies and we couldn’t have achieved the same success without them. We’re excited for the future of deltaDNA and the tools they will create to boost our business!”
George Boychev, Product Manager, Casualino JSC