A/B Experiments
How can I target only specific subscribers with an A/B experiment?
How can I target only specific subscribers with an A/B experiment?
When creating the experiment, you can configure audience tags:
Use tags to include or identify the audience you want to analyze.
Loop also applies distinct tags for control and variant groups, so you can segment your subscribers in external tools by which variant they saw (for example, in Klaviyo or your data warehouse).
Make sure your cancellation flow conditions are configured so that only the intended audience actually reaches that benefits page or offer.
What’s the difference between A/B experiments and regular edits to my benefits/offers?
What’s the difference between A/B experiments and regular edits to my benefits/offers?
Regular edits: You change content for everyone and rely on before/after analysis or intuition.
A/B experiments: You show two versions at the same time to comparable audiences and measure:
Save rate for A vs B
Attempts and saves over time
Whether B is statistically better according to your winning criteria
Experiments are the recommended way to make data-backed decisions for high-impact steps like cancellation benefits and offers.
Why is Experiment [A/B] not complete yet or is in progress?
Why is Experiment [A/B] not complete yet or is in progress?
Experiment only gets auto-completed once the experiment crosses the experiment completion configuration - number of days or number of cancellation attempts, until then, Experiment shows in In Progress state.