Experiments & A/B Testing

  • Controlled Experiments

    Build queries to maintain tight control of the player pool from which the randomly selected experimental player groups will be selected.

  • Define Variants

    Precisely define the variant configurations that players put in an experiment will experience—give some players higher virtual goods costs, some lower costs and discover how this affects their IAPs behavior.

  • Experiment Results

    Review the results of the experiment to analyse the impact on player behavior when subjected to the variant configurations you’ve specified under the experiment.

Key features

Experiments Wizard

Use our experiments wizard to quickly set up your upcoming experiments for the experiment period, target player pool, variant configuration, and experimental measures. An intuitive UI at each stage makes a complex configuration process very accessible.

Target Players

Use precise query rules to determine just which players you want to target for an experiment—you may only want players who’ve reached a certain skill level in your game to form the pool of players from which players put into the experiment will be randomly chosen.

Interim Results

View the results so far of an experiment while it’s still running and start to analyse player behaviour for the variant configurations you’ve defined for an experiment. If players in the experiment are not responding, you can abort the experiment.

Who uses GameSparks?