# What is a Test/Variant Group

In Optibase, a test helps you figure out what works best on your website by comparing different versions of your content — called variants — and tracking which version leads to more conversions.

There are three main types of tests in Optibase:

* A/B Tests (Variant Groups)
* URL Split Tests
* Multivariate Tests

***

### Why it matters

Running tests lets you make data-informed decisions about your site design and content. Instead of guessing what works, you can test it — and improve conversions without blindly redesigning pages.

***

### Test Types in Optibase

#### 1. A/B Test (Variant Group)

An A/B Test, also called a Variant Group, is where you test one specific change.

You start with a base version (your control), then create one or more variants with a small change:

* A different button color
* New headline text
* Swapped image or icon
* Tweaked form layout

When visitors land on your site, each one is randomly shown one of the variants. Optibase tracks how each performs based on the conversion goal you set.

> *Example: You test a green “Buy Now” button vs. a blue one to see which drives more purchases.*

***

#### 2. URL Split Test

A URL Split Test lets you compare entirely different pages.

Each variant is a separate Webflow page (with its own URL), and you direct traffic between them to compare performance.

This is useful when:

* Testing different layouts or flows
* Comparing two landing pages
* Validating a full-page redesign

> *Example: You test a long-form sales page vs. a short, punchy version to see which one converts better.*

***

#### 3. Multivariate Test

A Multivariate Test allows you to test multiple elements on the same page at once — for example, testing a headline, an image, and a button together.

Each element you want to test is placed in its own Variant Group, which works like a mini A/B Test.

Optibase then creates combinations of all possible variants and measures:

* How each Variant Group performs on its own
* How combinations of changes perform together

> *Example:*\
> *Headline A + Button B + Image C vs.*\
> *Headline B + Button A + Image C, etc.*

This helps you understand not just individual changes, but how they interact.

***

### Additional Testing Options

Optibase lets you control exactly who sees your tests by setting audience restrictions. You can limit exposure based on:

* Location (Geo-Restriction)

  Target visitors from specific countries or regions. Ideal for localized content or region-specific offers.
* Screen Size

  Restrict tests to certain viewport widths — for example, only users on mobile screens or large desktops. Helpful for responsive design changes.
* Operating System

  Show tests only to users on a specific OS like macOS, Windows, iOS, or Android. Great for testing platform-specific behavior or styling.
* Browser

  Limit tests to users on specific browsers such as Chrome, Safari, Firefox, or Edge. Useful when debugging or optimizing for browser quirks.<br>

> *Example: Want to test a new layout for desktop users on macOS using Safari? You can target that exact audience.*

These filters help you run more precise and relevant experiments — especially when testing features that behave differently across environments.

***

### Tips for Success

* Use A/B Tests for small, focused changes.
* Use Split Tests when comparing entire layouts or flows.
* Use Multivariate Tests when testing multiple elements at once — but keep combinations manageable.
* Set clear conversion goals before starting a test.
* Avoid testing too many things at once unless you have high traffic volume.