search

Test Types

Optibase offers three primary testing methods to help you optimize your conversion rate. Choosing the right one depends on the scale of your changes and the volume of traffic your site receives.

hashtag
1. A/B Test (Variant Groups)

The most popular choice for testing specific elements on a single page.

An A/B test (or A/B/n test) is used when you want to compare different versions of specific elements without changing the entire page.

  • When to use: Testing button colors, headlines, hero images, or CTA copy.

  • How it works: You keep your page structure identical but swap out specific components for different users.

  • Best Practice: Only run one A/B test per page at a time. If you run multiple independent A/B tests on the same page, your data may become skewed because the system cannot account for how the different tests interact with each other.

hashtag
2. URL Split Test

Best for comparing entirely different layouts or user journeys.

A URL Split Test (or Redirect Test) splits your traffic between two or more separate URLs at a percentage you define (e.g., 50/50).

  • When to use: Testing a complete homepage redesign, comparing a long-form landing page vs. a short-form version, or testing entirely different checkout flows.

  • How it works: Optibase redirects a portion of your traffic from the "Control" URL to a "Challenger" URL.

  • Advantage: It is ideal for "radical redesigns" where the changes are too complex to implement as simple element swaps on a single page.

hashtag
3. Multivariate Test (MVT)

"A/B Testing on steroids"—for advanced users with high traffic.

A Multivariate test allows you to test multiple elements on the same page simultaneously to see which combination of changes performs best.

  • When to use: When you want to test how your Headline, Hero Image, and CTA button interact with each other.

  • How it works: You create "Variant Groups" for each element. Optibase then automatically generates every possible combination (e.g., Headline 1 + Image 2 + CTA 1) and tracks which specific pairing drives the most conversions.

  • Traffic Requirement: Because MVT creates many combinations, you need significantly more traffic than a standard A/B test to reach statistical significance.

PreviousTestsNextWhat is a Tested User

Last updated