A/B Test - Multivariate Experiments directly in Shopware

by swkweb

Rating: no rating ( write a review )

Downloads: <10

Set up and run A/B tests directly in the Shopware administration to optimise your store based on real data and statistics, not just assumptions or gut feeling.

Annually €50.00* /month €659.40 * €599.95* /year 9.02% discount In comparison to the regular monthly price

Monthly €54.95* /month Cancelable monthly

Compatible with:

6.6.0.0 - 6.7.16.0

Includes:

Free trial month

The subscription starts with a free trial month during which it can be canceled at any time.

All updates

During the subscription all released updates are included for free

Support from the Extension Partner

In case you have any questions or problems with your extension you can always ask the Extension Partner for help during the subscription.

Description

Highlights

Make data-driven decisions instead of relying on gut feeling – sustainably increase your conversion rate
Run A/B and multivariate tests with any number of variations directly in your Shopware 6 store
Intuitive management of all experiments and variations within a dedicated area of the Shopware administration
Adjust products, categories and landing pages without coding skills – only basic knowledge is required for template (Twig) or style (CSS) adjustments
Full flexibility for developers through Twig conditionals and PHP integration for advanced customisation

Features

Create and manage A/B test experiments directly within the Shopware administration
Native integration with the Shopware Rule Builder: select a variation as a condition and leverage all rule-based Shopware features for A/B testing
Add as many variations per experiment as needed and label them individually for better structure
Define custom traffic distribution percentages for your variations – evenly or freely configurable
Customise products, categories, landing pages, SCSS/CSS, and Twig templates per variation through a modular element setup

Run several experiments on a sales channel at once – for the entire storefront, individual landing pages or Rule Builder conditions
Test any shop area flexibly using Twig if-conditions and PHP access for agencies and developers
Easily activate or deactivate entire experiments or individual variations with a single click
Automatic data transmission to Google Analytics (experience_impression event) for evaluating test results

Show more Show less

About the Extension

Make data-driven decisions instead of guessing

Which headline performs better? Which layout boosts conversions? How should you present specific products or categories? Until now, these decisions were often based on intuition, but that time is over.

A/B testing for smarter business decisions

With our A/B Test plugin, you make business decisions based on real data. Create experiments with as many variations as you need and define how traffic should be distributed among them. Test products, categories, landing pages, styling (SCSS/CSS), or entire Twig templates, all configurable directly from your Shopware administration. Several experiments can run at the same time.

For agencies and developers, the plugin offers maximum flexibility:
Access the active variation at any point within templates or the PHP stack to customise behaviour. This allows you to adjust the checkout, account area, header, footer, or any other component to fully leverage the potential of A/B testing.

Evaluate your experiments

For simple order evaluation, a built-in analytics dashboard is available. For more complex evaluation, you can use Google Analytics: If your shop is connected to Google Analytics, the plugin automatically sends the experience_impression event as part of the default Google Analytics integration.
See the official documentation for details:
https://developers.google.com/analytics/devguides/collection/ga4/integration

You can also evaluate test results with other tracking or analytics tools by passing the corresponding data manually.

Boost your conversion rate through A/B testing

We’re confident that A/B testing will help improve both user experience and revenue in any Shopware store. Try it out, and feel free to reach out if you have any questions!

Questions? Get in touch

Do you have an idea for an A/B test but are unsure how to implement it?
Feel free to contact us — we’ll be happy to advise you!
We also accept requests for custom implementations and new feature development.

Email: support@swk-web.com
Phone: +49 176 8057 9091

This extension is highlighted as part of the Premium Partner Program and can therefore be displayed with priority in the store.

Details

Available: English, German
Latest update: 4 November 2026
Publication date: 31 May 2026
Version: 3.0.0
Category: Conversion Optimization

Resources

Reviews (0)

Write a review

If you downloaded this extension you can write a review in the Shopware Account.

Write a review

No reviews found.

About the Extension Partner

swkweb

See all extensions

Partner Status

Shopware
Premium Extension Partner

Details

Ø-Rating: 5

Average rating of 5 out of 5 stars
Partner since: 2015
Extensions: 45

Support

Based in: Germany
Speaks: German, English

tl;dr

Navigate to "Content -> Experiments" in the administration
Create a new experiment
Define one or more variations for the experiment and assign them percentage shares for distribution
Create the elements (product, category, landing page, SCSS, Twig) you need for your test for each variation
Start the experiment by assigning it to a sales channel
Evaluate the results after some time via the built-in analytics dashboard or Google Analytics

Create experiment

Go to Content -> Experiments in the administration
Click on New experiment
Give the experiment a name
Save the experiment

Create variations

Create a new variation in the experiment via the Create variation button
Give the variation a name and assign it a colour
Set the percentage share with which the variation should be distributed
Save the variation

Customise variation elements

Click Create element within a variation and select the desired type (Product, Category, Landing page, SCSS/CSS, Twig/HTML)
Make the desired adjustments for the element (see below)
Save the element

SCSS/CSS elements

Give the element a name
Activate the element
Define the desired CSS styles in the editor

Twig/HTML elements

Give the element a name
Activate the element
Select the file path and block name to be overwritten
Define the desired HTML/Twig in the editor

Product elements

Select the desired product
Activate the element
Select the languages for which the adjustments should apply
Customise product name, description, custom fields and layout as needed

Category elements

Select the desired category
Activate the element
Select the languages for which the adjustments should apply
Customise category name, description, custom fields and layout as needed

Landing page elements

Select the desired landing page
Activate the element
Select the languages for which the adjustments should apply
Customise landing page name, meta title, meta description, custom fields and layout as needed

Start experiment

Open the Overview tab of the experiment
Under Sales channels & scope, select the sales channels the experiment should run on. This is possible once at least one variation receives a share of the visitors
Save the experiment

Running several experiments at once

Several experiments can run on a sales channel at the same time
Each experiment shows its scope under Applies to, which follows from the elements of its variations:
- Entire storefront: SCSS, Twig, product and category elements. A visitor takes part in at most one of these experiments, so they cannot override each other
- Targeted landing pages: only landing page elements. Visitors are assigned to a variation when they first open one of the landing pages, which are not cached while the experiment runs. A landing page can only be part of one running experiment per sales channel
- Rule Builder conditions only: no elements. The experiment takes effect through the "A/B test variation" rule condition, for example in promotions or shipping methods
Switch on Apply to the entire storefront when a rule changes product prices with a variation of the experiment, for example through advanced prices, or when your own Twig templates check the visitor's variation (see Tips for developers). Visitors are then assigned before the HTTP cache and see their variation's prices and templates on every page, and the experiment falls under the Entire storefront scope. As long as the experiment does not apply to the entire storefront, its form warns about such price rules
Don't change whether a running experiment applies to the entire storefront: that changes who takes part and skews its results. The administration warns before you save such a change. To start afresh, remove the experiment from its sales channels and create a new one

Tips for developers

In Twig templates, the global parameter swkwebAbTestContext holds the visitor's variation. Define conditions with it, e.g. {% if swkwebAbTestContext.variationTechnicalName == 'my-technical-name' %}
swkwebAbTestContext is only set when the visitor sees a variation of an experiment for the entire storefront. Landing page and rule experiments are not part of the HTTP cache key, so templates must not depend on them
swkwebAbTestContext is not set for the original group either, since it shares its cache entries with visitors outside every experiment. So only check for the variations in your templates and handle the original group in the else branch
An experiment that only changes your own templates has no elements, so it does not apply to the entire storefront. Switch on Apply to the entire storefront for it, so your templates can check the visitor's variation
In the PHP stack you can access the visitor's A/B test context through the request: $abTestContext = \Swkweb\AbTest\Core\AbTestRequest::context($request);
$abTestContext->getAssignments() returns every experiment the visitor takes part in

Evaluate experiments

For simple order evaluation, a built-in analytics dashboard is available
Experiments and variations in which orders have been placed can't be deleted, so the orders stay assigned to them. Archive an experiment you no longer need instead, and set a variation's share to 0 to keep new visitors from getting it
For more complex evaluation, you can use Google Analytics
By default, every experiment a visitor takes part in is reported to Google Analytics once per browser tab, as an experience_impression event with the exp_variant_string parameter swkwebAbTest-- (original for the original group)
This allows you to evaluate test results and create reports in Google Analytics

3.0.0 6.7.0.0 - 6.7.16.0 4 November 2026

Feature: Add landing page elements to override a landing page's name, meta title, meta description, custom fields and layout
Feature: Run several experiments on a sales channel at once, each applying to the entire storefront, to targeted landing pages or only through Rule Builder conditions, depending on its elements or the new "Apply to the entire storefront" switch (see "Running several experiments at once" in the manual); existing experiments keep applying to the entire storefront
Feature: Report a visitor's experiments to Google Analytics once per browser tab instead of on every page view; existing reports therefore show fewer experience_impression events from this version on
Feature: Allow starting and stopping experiments with the "Edit" permission alone; existing roles are updated automatically
Feature: Add keyboard shortcuts, a warning about unsaved changes and error markers to the experiment editor
Feature: Set the shares of all variations directly in the variation overview
Feature: Preselect a colour for new variations
Bugfix: Stop ignoring the customised layout slots of product and category elements in Shopware 6.7.10+, and in earlier versions those of category elements in languages other than the system language
Bugfix: Stop deleting the experiment's last variation along with the selected one
Bugfix: Show the loading state while an element's layout and custom field sets load
Bugfix: Stop the analytics from showing the previously opened experiment
Bugfix: Follow the Administration's dark mode in the analytics
Bugfix: Point out empty required fields when saving an experiment fails instead of reporting success
Bugfix: Stop deleting saved elements when leaving them with an empty required field
Performance: Landing pages targeted by a running landing page experiment are not stored in the HTTP cache, and while a rule experiment runs, a visitor's first uncached response is not stored either, so the cache hit rate for first-time visitors drops
Breaking Change: The "A/B test variation" rule condition now checks all of a visitor's experiments, and "is none of" also matches visitors outside the experiment, including every customer while no experiment runs; check the promotions, prices and other rules using "is none of" before updating, as they may otherwise apply to all customers
Breaking Change: Record an order's experiments in the swkwebAbTestAssignments association instead of the swkwebAbTestVariation custom field; existing orders are carried over, but exports, flows and integrations reading the custom field receive no data for new orders
Breaking Change: Prevent deleting experiments and variations that orders were placed in; archive them instead
Breaking Change: Remove window.swkwebAbTestContext and the storefront/component/analytics.html.twig extension; tracking scripts can read a visitor's experiments from the swkweb-ab-test-variation cookie instead, which is now readable via JavaScript and holds .-separated - pairs (cookies from earlier versions remain valid)
Breaking Change: Remove the swkwebAbTestVariation Twig parameter; swkwebAbTestContext is now only set for visitors seeing a variation of an experiment for the entire storefront; custom templates still using swkwebAbTestVariation show the original to every visitor without an error, so check them before updating
Breaking Change: Replace the sales channel's swkwebAbTestExperimentId field and swkwebAbTestExperiment association with the swkwebAbTestExperiments association; templates and integrations still using the old ones get no value without an error
Breaking Change: Remove AbTestContext::isValid() and getRequestedVariationId(); getExperimentId(), getVariationId() and getVariationTechnicalName() now only describe the experiment for the entire storefront, while the new getAssignments() and getAssignment() cover all of the visitor's experiments

2.1.0 6.7.0.0 - 6.7.16.0 22 September 2026

Feature: Let experiments be archived
Feature: Add a sidebar filter panel to the experiment listing
Feature: Support Shopware's reworked HTTP cache (CACHE_REWORK feature flag)
Feature: Rework validation of variation technical name and percentage
Feature: Default analytics start date to the experiment creation date
Feature: Add admin ACL roles
Bugfix: Resolve the variation from the current request's sales channel
Bugfix: Recompile the storefront after a variation's Twig overrides change
Bugfix: Fingerprint compiled variation stylesheets
Bugfix: Vary the 404 error page cache by variation
Bugfix: Invalidate the HTTP cache when an experiment changes
Bugfix: Make the activation route agree with the cache hash it emits
Bugfix: Vary the ESI header and footer cache by variation
Bugfix: Properly size administration buttons
Bugfix: Pin the decoration order of the HTTP kernel
Bugfix: Keep the reworked cache from storing a cleared variation cookie
Bugfix: Honour fractional variation percentages
Bugfix: Negate "hide in navigation" when overriding a category
Bugfix: Stop category elements from activating their category
Bugfix: Reset the Twig namespace hierarchy between requests
Bugfix: Expose the variant technical name in the analytics response
Bugfix: Resolve the variation with a GET sub-request
Bugfix: Invalidate the variation cache when a variation is inserted
Bugfix: Use the indexed order_date column for the analytics date filter

2.0.1 6.7.0.0 - 6.7.16.0 22 July 2026

Feature: Add time-of-day selection to analytics date filters
Bugfix: Do not resolve variations if the request attribute is already present
Bugfix: Clear attributes of variation resolution subrequests
Bugfix: Make assignments to the original group experiment-scoped
Bugfix: Link orders from the original group to their experiment and correct the target-percentage calculation
Bugfix: Correct percentage-based distribution of variations

2.0.0 6.7.0.0 - 6.7.16.0 31 May 2026

Compatibility: Shopware 6.7

1.1.1 6.6.0.0 - 6.6.10.31 20 July 2026

Feature: Add time-of-day selection to analytics date filters
Bugfix: Do not resolve variations if the request attribute is already present
Bugfix: Clear attributes of variation resolution subrequests
Bugfix: Make assignments to the original group experiment-scoped
Bugfix: Link orders from the original group to their experiment and correct the target-percentage calculation
Bugfix: Correct percentage-based distribution of variations

1.1.0 6.6.0.0 - 6.6.10.31 31 May 2026

Feature: Add possibility to open variation in the storefront
Bugfix: Make variation color field required
Bugfix: Properly consider variation in the route cache

1.0.0 6.6.0.0 - 6.6.10.31 31 May 2026

First Shopware 6 release