Experiments: A/B Tests and Personalizations in Pack
An experiment changes what some of your visitors see. Pack has two types, and they answer different questions.
| A/B test | Personalization | |
|---|---|---|
| The question | Which version of this content performs better? | What should this group of visitors see? |
| Variants | A control and one or more variants | One variant |
| Who sees it | Traffic is split across the variants | Everyone in the audience sees the variant; everyone else sees your default content |
| Audience | Optional | Required |
| Reporting | Conversions per variant, with lift and chance to win | Exposures; measure conversions in your analytics tool |
The two work well together: once an A/B test shows what works for a segment, a personalization can keep showing it to that segment.
Personalization is not self-serve yet. Contact Pack to have it enabled for your storefront. It requires A/B Testing to be enabled first.
Where experiments live in the admin
Once Pack enables personalization for your storefront, A/B tests in the left sidebar becomes Experiments, and Audiences appears below it. Both appear for storefronts hosted on Oxygen. Stores without personalization keep the A/B tests page exactly as it was.
The Experiments list shows both types together, with a Type badge on each row. It is ordered by status, running first, then scheduled, paused, draft and ended, and most recently updated first within each.
- Use the All, A/B Tests and Personalization filters at the top to narrow the list.
- Use the search box to find an experiment by its name, handle or description, a variant's name, or the name of its audience.
The filter and the search are kept in the page's URL, so you can bookmark or share a filtered list.
The … menu on each row ends or deletes the experiment. An experiment can only be deleted once it is no longer running or paused.
Creating an experiment
- Go to Experiments in the left sidebar
- Click New experiment
- Choose A/B Test or Personalization
Each choice opens that type's own form:
Where to go next
- A/B Testing — variants, running, scheduling and promoting a winner
- Personalization — the variant, the audience, running and reporting
- Audiences — the reusable targeting both types use