Leep.
HomeBusiness DevelopmentWeb & EngineeringWork
Contact
Log in
Leep.

Custom websites and web applications for ambitious companies.

Navigation

  • Home
  • Work
  • Contact

Services

  • Website
  • App
  • AI Implementation

Resources

  • Articles
  • Dictionary

Contact

  • hello@leep.no

© 2026 Leep. All rights reserved.

Back to Dictionary
Marketing

What is A/B Testing?

Last updated: January 15, 2025

On this page

TL;DRExampleExplanationWhy It MattersRelated Terms

TL;DR

A/B testing (split testing) means showing two different versions of something to different users to see which performs better.

Example

Should your "Buy Now" button be green or blue?

Instead of guessing, you test it:

  • 50% of visitors see a green button (Version A)
  • 50% of visitors see a blue button (Version B)
  • After 1,000 visitors, you compare which button got more clicks

If the green button got 3% clicks and the blue button got 4.5%, blue wins. Now you know, based on data, not opinion.

What you can A/B test:

  • Headlines and copy
  • Button colors, text, and placement
  • Images and videos
  • Page layouts
  • Pricing displays
  • Email subject lines
  • Form length and fields
  • Navigation menus

Explanation

How A/B Testing Works

  1. Hypothesis - "I think changing X will improve Y"
  2. Create variations - Make version A (control) and version B (variant)
  3. Split traffic - Show each version to a random subset of visitors
  4. Measure results - Track conversions, clicks, or other goals
  5. Analyze - Determine if the difference is statistically significant
  6. Implement winner - Roll out the winning version to everyone

Statistical Significance

The hardest part of A/B testing is knowing when you have enough data to trust the results.

If Version A has 2.1% conversion and Version B has 2.3%, is B really better? Or did you just get lucky with that batch of visitors?

Statistical significance (usually 95%) means you can be confident the difference is real, not random chance.

Common A/B Testing Mistakes

  • Ending tests too early - Not enough data to be sure
  • Testing too many things at once - You won't know what caused the change
  • Testing tiny changes - Small tweaks often don't have measurable impact
  • Ignoring segments - The winner overall might not be the winner for mobile users

Why It Matters

For Business Owners

Stop arguing about opinions. "I think blue is better" vs "I think green is better" goes nowhere. A/B testing provides data to settle the debate.

Compound improvements. A 10% improvement from one test, plus 10% from another, plus 10% from another... compounds into significant gains.

Reduce risk. Before making a big change, test it on a portion of your traffic first.

When to A/B Test

A/B testing requires enough traffic to get statistically significant results. If you have 100 visitors per month, you probably can't run meaningful tests.

As a rough guide: you need hundreds of conversions to test most things reliably.

Related Terms

UX

UX (User Experience) is how a person feels when using a product, website, or app. Good UX means the experience is easy, intuitive, and enjoyable.

KPI

A KPI (Key Performance Indicator) is a measurable value that shows how effectively you're achieving your business objectives.

Conversion

A conversion is when a visitor takes a desired action: buying a product, signing up for a newsletter, filling out a contact form, or any other goal you define.

Analytics

Analytics is the collection and analysis of data about your website or app to understand user behavior and business performance.

Need help with your digital project?

We build websites, apps, and digital solutions for businesses.

Get in touch