A/B Testing Campaigns

the new a/b testing feature is in beta testing this feature will only be available to marketing pro+ subscriptions campaigns are at the center of your marketing journey send them via various chanab testing allows you to compare versions of your marketing campaigns against one another to determine the optimal message to send out you'll be able to use data to drive your decision making instead of intuition prior to creating an a/b test or reading through this document, we recommend reading through campaign builder docid\ yeopwsehov3mfydjilqzo first create an a/b test navigate to marketing > campaigns and click add (+) on the channels screen, under "campaign type", click "a/b test" under "campaign channel", select either email or text before hitting next fill out "recipients" as needed next, select how many "variants" and what percentage of the audience you want allocated for each variant on the "content" step, include the desired content upon hitting next, you will be brought to fill out the next variant this step will repeat until all variants have some type of content included you can switch between variants by clicking on the tabs above the content design finally, you'll see the "review" step hit "launch campaign" a/b testing faqs which channels can i use a/b testing with? currently, users will be able to send a/b tests via email or text other channels are not permitted at this time if you create an a/b testing campaign, you will also not be able to create a waterfall or co marketing campaign how many variants can i create? currently you can create up to 3 by default, the flow will include 2 groups with 50% of the audience allocated to each group you are able to customize the percentage allocation but recommended best practices would be to have an equal allocation for each group can i test multiple variables at once? yes there is no limitation on what changes you choose to make from one group's content to another, outside of it needing to be in the same channel however, recommended best practices for a/b testing include testing only one change at a time , so that you can analyze and determine what specific change is improving your campaign outcomes how large should my audience be for the test? use a sample size calculator to determine the minimum size needed for your test typically, most people default to a 95% confidence interval when determining the size of their test what if i'm using dynamic content? if you are using dynamic content that will result in uneven a/b testing variants, aiq will prioritize ensuring there is the correct split in the audience rather than the largest possible audience that can be reached do you support automated win / race conditions? no