A/B testing a homepage with little traffic

Key takeaways

A homepage A/B test needs far more traffic than most small business sites have to detect a realistic improvement with confidence. With a few thousand homepage visits a month, test only big, bold changes, or use before-and-after comparisons over four-to-eight-week periods with a change log, user tests and a careful look at seasonality.

• Small sites rarely have enough traffic for reliable A/B tests.
• Test only big changes, such as a new headline and hero.
• Otherwise compare 4-to-8-week periods with a change log.
• Use 5-person user tests to check changes before launch.

Can a small business A/B test its homepage?

Sometimes, but rarely in the way large sites do. An A/B test splits visitors between two versions and compares results. Detecting a modest improvement in enquiry rate with confidence needs thousands of visitors per version, and many small business homepages get a few hundred enquiry-ready visits a month.

• Why your homepage is not converting visitors into customers
• A/B testing for small business websites

When is an A/B test worth running?

Run one when the homepage gets enough traffic, the change is big enough to make a large difference and you can wait several weeks. A completely different headline and hero is testable; a button colour change is not, at small-business traffic levels.

• How much traffic you need to trust a website test

What should you do instead?

Combine three lighter methods.

• Before and after: compare four-to-eight-week periods, with the change logged.
• User tests: show the new version to five target buyers before launch.
• Leading signals: watch next-step clicks, which move faster than enquiries.

• Change one thing at a time: a homepage change log
• User testing for small business websites
• How long until homepage changes show results?

How do you run a test without harming SEO?

Google's guidance on testing is to avoid cloaking, use rel=canonical on variant URLs pointing to the original, use 302 redirects for redirect tests and end the test promptly. Most testing tools handle this, but check before running a test on your most important page.

• Homepage SEO vs conversion: write for both

When should you not test at all?

Do not test when the current homepage has an obvious flaw, such as no clear headline or a broken form. Fix it. Testing is for choosing between two reasonable options, not for proving that a broken page is broken.

• Which homepage problem should you fix first?

What does this look like in practice?

A pattern we see: a firm runs a button-colour test for three months and gets no clear answer, because each version received a handful of enquiries. Rewriting the headline, logging the change and comparing the next six weeks with the previous six usually gives a clearer, quicker read.

• See client work
• Website Design & Development for UK businesses

Homepage testing checklist

• Check monthly homepage sessions and key events.
• Run A/B tests only for big changes with enough traffic.
• Otherwise log changes and compare 4-to-8-week periods.
• Run a 5-person user test before launch.
• Follow Google's guidance to protect SEO during tests.

• A homepage dashboard: five numbers to check each month
• What Search Console says about your homepage

Next step

Note your monthly homepage sessions and enquiries. If enquiries are under a few dozen a month, plan before-and-after comparisons rather than A/B tests.