A/B testing - the video

Test instead of guessing. The video shows how to set up an experiment with shopping experience layouts in Shopware and when a result really counts.

To the product page To the manual
01
Walkthrough

What the video shows.

  1. A new experiment takes a few steps. Whole shopping experience layouts are tested, variant B gets a different layout.
  2. As many variants as you like are possible, each with its own traffic share.
  3. A result only counts once sample size and confidence level are reached. Before launch the plugin estimates whether your traffic is enough and calculates the sample size per variant.
  4. Bots are not counted, assignment runs server-side without flicker.
  5. It measures conversion and revenue per visitor, not just clicks.
02
At a glance

The essentials in brief.

Layouts

as variants in the test

Traffic share

per variant, freely set

Sample size

calculated before launch

Revenue

per visitor, not just clicks

03
FAQ

Questions about the video.

What exactly is tested?

Whole shopping experience layouts. Every variant gets its own layout and its own traffic share.

When is a result meaningful?

Only once sample size and confidence level are reached. Before launch the plugin calculates how much traffic each variant needs.

Does the page flicker when a variant is shown?

No. Assignment runs server-side without flicker, and bots are not counted.

04
Read on

Related pages.

05
More videos

More on video.

All videos →