What is baseline reporting in Experiments and what does it mean for me?
As part of our drive to constantly bring innovation to the Intelligent Reach Platform, we have introduced new and more advanced reporting methodology for our Beta Experiment customers.
You will see these new stats available for all A/B experiments (past and present) when you navigate to the Experiments module:
The baseline methodology has been introduced as a means of further removing the impact of external influences on Experiments running through the Platform.
This analytical model does this by directly comparing the relative increase from within each experiment set, mitigating the impact of key influences such as an imbalance caused by the presence of Best Selling products in one of the Experiment sets.
What this means, is that this model should be an even more accurate representation of the Experiment outcome, but we encourage you to review both the Head 2 Head and the Baseline reporting methodologies together when reviewing your experiment results.
If you wish to continue to report using the previous statistical model, this is still available from within the Platform by switching the reporting view between 'Baseline' and 'Head 2 Head'.