GDP is a revenue optimization company. Our core thesis is that revenue optimization in gaming is best done via personalization, and that personalization is best done by using advanced data science (both the more traditional machine-learning approaches and the recent AI-based approaches).

This naturally implies that we run a lot of experiments.

Done properly, experimentation is an essential part of any revenue optimization practice. With the inclusion of rigorous data gathering and experimentation, revenue optimization is an empirical science. And without them, revenue optimization is just informed speculation.

GDP Principal Scientist Julian Runge has two recent publications on experimentation.

• The first, Gaming companies run thousands of experiments a year – here’s why and how, is a comprehensive overview article by Julian on the relationship between the different stages of a game’s lifecycle and the types of experiments that are appropriate for the game.
• The second, “Is Gaming Better Than Everyone at Experimentation?“, is an appearance on the Game Economist Podcast, where Julian expands on the article and answers famed game economist Phillip Black’s drill-down questions.

If you’re looking for a deeper understanding of the role of experimentation in modern gaming, we highly recommend both reading the article and listening to the podcast.