Emarsys offers a built-in A/B testing feature available through all delivery channels. A/B tests lets you compare the performance of your existing promotional solutions (website promotional boxes, static banner ads, newsletter promotions) with the performance of our personalization solutions.
We currently do not offer an interface for A/B testing recommendations in email campaigns. For more information on how to conduct A/B tests on Emarsys email campaigns, see Versions tab.
A/B tests track performance throughout your online operations, across all delivery channels, so you can track each banner or newsletter click down to the actual conversions generated by them.
In the standard A/B test setup, your existing promotional solutions is labeled as BASELINE, while our recommender solution is labeled SCARAB. We use your existing delivery integration to receive visitor requests, then split visitors randomly between the BASELINE and SCARAB groups, so 50% of your visitors see your existing solution, and 50% receive personalized recommendations.
The BASELINE values are passed on to our system using the recommender requests themselves. The integration is different for each delivery channel, but the logic is the same.
Use the show A/B tests option in the Dashboard Statistics to display A/B test results.
Setting up A/B tests for website recommendations
Setting up the A/B test requires you to pass the baseline setting to the “recommend” command, which is part of the snippet you use to request the recommender widget.
<div id="my-recommendation-container"></div>
<script>
ScarabQueue.push(["recommend", {
logic: "PERSONAL",
containerId: "my-recommendation-container",
limit: 3,
baseline: ['item_4', 'item_9', 'item_7']
}]);
</script>Once your baseline is setup, please contact our Customer Service to actually enable and bring the A/B test live.