What Is A/B Testing and How to Use It on Your Site and Ads

A/B testing compares two versions of the same thing to see which gives a better result. Instead of guessing whether “Free property valuation” or “Sell your flat faster” is the better headline, you run both and measure which brings more enquiries.

How an A/B test works

1. Choose one thing to change, for example a headline, button or image.
2. Half the visitors see version A and half see version B.
3. Measure how many conversions each version brings.
4. When the difference is large and reliable enough, keep the better version.

Change one thing per test. If you change the headline, image and form at once, you will not know what made the difference.

What you can test

• Website: headline, button text, layout, form length, reviews, images.
• Ads: headlines, images, videos, offer.
• Emails: subject line, send time, call to action.

Testing is built into Meta ads: you run several creatives and the platform moves the budget towards those that perform better.

How much traffic you need

An A/B test on a website needs enough visits for the result not to be chance. A reliable result usually takes a few thousand visits and at least dozens of conversions per version.

Most small businesses do not have that much traffic on one page. A better approach then is to:

• make changes that have worked well on similar pages
• compare the period before and after the change
• watch heatmaps and session recordings to see whether behaviour changes

Common mistakes

• Stopping too early. One version leads on day one and the test is stopped. That result is often not real.
• Too little traffic. A difference of two enquiries is not proof.
• Too many changes at once. Nobody knows what worked.
• Measuring clicks instead of enquiries. The winner is the version that brings customers, not clicks.

I run A/B tests and page changes in the conversion rate optimization service, and creative testing in Meta Ads.