A/B Testing for Small Websites: What to Test First and How

A/B testing means showing two versions of a page or element to different visitors and measuring which performs better. Large companies run many tests at once. Small businesses can benefit too, but only if they approach testing realistically, with the right expectations about traffic, time and what is worth testing.

This guide explains how A/B testing works, when it makes sense for a small website and how to run tests that produce useful answers.

How A/B testing works

In a simple A/B test, visitors are randomly split into two groups. Group A sees the current version, often called the control. Group B sees a changed version, called the variant. You then compare a specific goal, such as form submissions, WhatsApp clicks or purchases.

The key ingredients are:

• One clear goal: the action you want more of
• One main change: so you know what caused any difference
• Random splitting: so both groups are similar
• Enough visitors and conversions: so the result is not just luck

Various testing tools can split traffic and track results. Some website builders and ecommerce platforms include basic testing features.

Do you have enough traffic?

This is the honest first question. A/B tests need a reasonable number of conversions in each version to give a reliable answer. If your page gets a handful of enquiries a month, a test could take a very long time, and small differences may be due to chance.

If your traffic is low, you can still improve your site by:

1. Fixing obvious usability problems found by watching real users
2. Following established good practice for forms, buttons and page structure
3. Testing bigger, bolder changes that are more likely to show a clear difference
4. Running tests on your highest-traffic pages or on paid ad landing pages where you control traffic
5. Using qualitative feedback, such as asking customers what nearly stopped them from contacting you

Many testing tools include a calculator that estimates how long a test needs to run based on your traffic and expected change. Use one before you start.

Key takeaway: on low-traffic sites, test big ideas, not tiny tweaks. Changing a button shade will rarely produce a measurable difference, but changing the whole offer or page structure might.

What to test first

Start where changes are most likely to matter. Good first tests for small business sites include:

• Headlines: a clear, specific headline versus a vague or clever one
• Offers: a free consultation versus a downloadable guide, or different trial offers
• Calls to action: different labels, such as "Get my quote" versus "Contact us"
• Form length: a short form versus a longer one
• Page structure: pricing shown upfront versus pricing further down the page
• Proof: with and without testimonials near the call to action

Focus on pages with clear goals and decent traffic, such as your main service page, a campaign landing page or your checkout. Our post on call-to-action placement offers ideas worth testing.

Running a fair test

Poorly run tests give misleading answers. To keep results trustworthy:

• Decide the goal and the minimum run time before starting
• Run both versions at the same time, not one week after another, because seasons, holidays and promotions affect behaviour
• Avoid stopping the test early just because one version looks ahead after a few days
• Do not change other things on the page during the test
• Run tests for at least a full week or two, so weekday and weekend behaviour are both included

Be aware of major events that affect traffic, such as festive seasons, sales campaigns or a sudden burst of social media attention. They can skew results.

Reading results and acting on them

When the test ends, check whether the difference is large enough and reliable enough. Most testing tools show a confidence measure. If the result is unclear, that is still useful information: the change probably did not matter much, and you can move on to a bigger idea.

Record every test in a simple log: what you changed, why, how long it ran and what happened. Over time, this log becomes a valuable record of what works for your customers. Apply winning changes permanently, and use what you learned to inspire the next test. Our article on forms people actually complete has plenty of form ideas to test.

FAQ

How long should an A/B test run?

Until it reaches the sample size estimated before starting, and usually at least one to two full weeks to capture different days of the week.

Can I test more than two versions?

Yes, but each extra version needs more traffic. Small sites are usually better off testing two versions at a time.

Does A/B testing affect SEO?

Testing tools used properly, with both versions showing similar content, generally do not cause problems. Avoid showing search engines different content from visitors.

A/B testing is a powerful way to learn what your customers respond to, but it rewards patience and realism. Check your traffic, test meaningful changes, run fair tests and record what you learn. Even a few well-run tests a year can steadily improve your website's results.