Experiments
Make any changes to your website using a simple visual editor.
Experiments let you make changes to your website using a visual editor — no coding or deployments needed. You can add or remove content blocks, insert weblayers, adjust colors and fonts, and personalize your website using customer data with Jinja. Like weblayers, experiments support multiple variants, audience targeting, and A/B testing.
Every time an experiment is applied to your website, it tracks a single event called experiment. You can use this data for further analysis in Bloomreach.
Why use experiments
Experiments let you make small, impactful changes to your website without writing code or involving your IT team. You can A/B test any change and evaluate results directly in Bloomreach. Customer data is available throughout, so you can personalize the web experience using Jinja.
How experiments work
An experiment is an ordered list of modifications. The order matters—modifications are applied to the webpage from top to bottom. Each experiment also has its own settings, just like any other campaign.
An experiment is a list of modifications. This list is ordered, so the order matters. The modifications are applied to the webpage in the exact order from top to bottom. There are multiple modification types to choose from, and they're described later in the section Modification types.
Each experiment also has settings associated with it, just like any other campaign. The settings are described in the section Settings.
Create an experiment
To create a new experiment, go to Campaigns > Experiments > New experiment

Navigating to New experiment from the Campaigns menu.
You'll be prompted to enter the URL of the page you want to modify. This can be your homepage or any other page. The URL you enter doesn't limit where the experiment runs. It's only used for setup.
Important
The page at the URL you enter must be integrated with the current project before you create the experiment. Otherwise, the experiment won't work. To preview experiments on pages that aren't integrated yet, use the Bloomreach Experiments Chrome extension.
After clicking Start, you'll be taken to your website and an inline visual editor (called the Experiments editor) loads.
Experiments editor
After clicking Start, your website opens with the Experiments editor loaded inline. Once the editor loads, you can select and modify any element on the page. For a full walkthrough of the editor interface, modification types, and selectors, see Use the Experiments editor.
Set up A/B testing
In the Design tab, you can create multiple variants and configure A/B testing. You can also update the preview URL, rename variants, and duplicate them.

The Design tab with variant controls and traffic distribution settings.
In the Design tab, you can create multiple variants of the experiment and set up A/B testing. You can also change the preview page URL, change the variant names, and duplicate them.
Traffic distribution controls how customers are split across variants:
- Automatic traffic distribution (default): Bloomreach identifies the best-performing variants and serves them more frequently.
- Custom traffic distribution: You define the percentage of customers who see each variant.
Both options support a control group — a segment of customers who continue to see the original page, used for more accurate evaluation.
Configure experiment settings
Adjust your experiment's settings in the Settings tab.

The Settings tab with experiment configuration options.
| Setting | Description |
|---|---|
| Audience | Define a custom audience using standard customer and event filters. |
| Category | Choose which consent is required for a customer to see the experiment. See Consent management for details. If you don't use consents, you'll see the campaign groups setting here. |
| Conversion goal | Drives automatic A/B test optimization and the generated evaluation report. |
| Display | Set how long to keep showing the experiment to each eligible customer: - Once: Shown once for the full duration of the experiment, tracked when experiment tracks action=show.- Once per visit: Shown once per session. Shown again only after the next session_start.- Always (default): Shown on every visit, every day, and after every page refresh. |
| JavaScript condition | Specify custom JavaScript as a display condition. This is an advanced feature. See Experiments for developers. |
| Schedule | Display the experiment immediately or set a specific time period. |
| Show on | Choose which URLs to show the experiment on. |
| Show on pages with variables | Display the experiment only on pages with specific pre-defined variables set up in your project settings. See Enhanced Web Targeting for details. |
| Conversion goal | The automatic A/B test optimization and the generated evaluation report are based on this setting. |
| Schedule | Choose to display the experiment instantly or set a specific time period in which the experiment is shown. |
| Show on | Choose on which URLs to show the experiment. |
| Show on pages with variables | You can choose to display the experiment only on pages with certain pre-defined variables that you have set up in the project settings. See Enhanced Web Targeting for details. |
| Target devices | Show on any device or only desktop or mobile. |
| Trigger | Choose On page load (default) or a specific Datalayer event. See Enhanced Web Targeting for details. |
Test your experiment
To test an experiment before launching, set the Show on condition to a URL containing a test parameter — for example, ?test. After launching, visit www.yourwebsite.com/?test to see the experiment in a real environment. You can share this link with others and it works until you stop the campaign.
Important
Experiments ignore URL fragments. Any
#character in a URL filter is ignored, and only the part before it is used.
Evaluate your experiment
Every experiment has a built-in evaluation dashboard showing common metrics and analyses. If A/B testing is enabled, the dashboard also shows how each variant is performing through funnels and accompanying tables.

The built-in evaluation dashboard showing variant performance and funnel metrics.
To customize the dashboard, save it as a custom dashboard using the menu in the top-right corner next to the Save button.
Integrate experiments on your website
Before going to production, read about how to integrate and use Experiments on your website correctly. This is a required step to optimize the experience for your users.
Related resources
If you need help with the editor or run into unexpected behavior, these pages have you covered:
- Use the experiments editor: Guide to the editor interface, modification types, variants, and selectors.
- Troubleshoot experiments: Solutions to the most common issues, from invisible experiments to CSS conflicts and editor errors.
Updated 4 days ago
