Split Testing a Small Business Website With Low Traffic

You changed the button colour, ran a split testing tool for two weeks, and got a result that flipped back and forth every few days. With 1,500 visitors a month, that is not bad luck. Most small business websites simply do not get enough traffic for classic split testing to give a reliable answer.

I design and rebuild websites for small firms, from dentists to joinery workshops, and most of them come to me with this exact frustration. The good news is that you can still test and improve conversions. You just need different methods, bigger changes and a little patience.

Why split testing struggles on small sites

A split test, often called an A/B test, shows half your visitors one version of a page and half another, then compares how many of each group convert. The maths that decides whether a difference is real depends on how many conversions you collect, not how many days pass.

Here is a rough sense of scale. Suppose your contact page converts at 3 percent and you want to detect a lift to 3.6 percent, a 20 percent improvement. Standard calculators suggest you need roughly 14,000 visitors per version, or around 28,000 in total, to be reasonably confident. At 1,500 visitors a month, that is more than a year and a half of testing for one change.

These are approximate figures, but the pattern holds. The less traffic you have, the bigger the change must be before a test can see it.

The counter-take: stop testing button colours

Popular conversion advice is full of stories about a red button beating a green one by 30 percent. Those stories usually come from sites with hundreds of thousands of visitors, and many of the results did not hold up when tested again.

On a small site, a colour tweak will never produce a measurable difference within a sensible time. You will either wait forever or, worse, stop the test early on a lucky week and roll out a change that does nothing. I would rather a client spent that energy on changes large enough to move behaviour: a new offer, a shorter form, a different headline that speaks to a different problem, or a clearer price.

If you want to improve your buttons, rewrite the words instead of the colour. Our guide on writing button copy that tells people what happens next has examples that make a real difference.

How to run split testing with low traffic

If you still want a proper test, these rules make it far more likely you get a usable answer.

Test one page with the most traffic

Pick the page that receives the most visits and leads directly to an enquiry, often the homepage or a main service page. Splitting already thin traffic across several tests at once guarantees none of them finishes.

Make bold changes

Test two genuinely different versions. For example, a page that leads with price and a booking form against one that leads with case studies and a phone number. Big differences produce big effects, which a small sample can detect.

Measure a step closer to the visitor

If you only get 15 enquiries a month, measure an earlier action as well: clicks on the main button, scrolls past the pricing section, or starts on the form. These happen more often and give you a signal sooner, as long as you check that they still lead to real enquiries.

Decide the length in advance

Set a fixed period, at least four full weeks so each weekday is covered several times, and do not peek and stop early. Write the end date down before you start.

Keep everything else steady

On a small site, outside events can swamp the result. A local newspaper mention, a paid ads push or a bank holiday week can double enquiries for a few days and land unevenly between the two versions. While a test runs, avoid launching campaigns, changing prices or editing other key pages. Keep a short log of anything unusual, such as a supplier email blast or a radio spot, with the date. When you read the result, check the log first. If a large share of the conversions arrived in one unusual week, treat the outcome as a hint to retest rather than a winner to roll out. It is also worth checking that both versions load equally fast on a phone, because a slower variant can lose for reasons that have nothing to do with the design.

Better methods than split testing for small sites

For most small businesses, these methods give more useful insight per hour than any split test:

• Five second tests and hallway tests. Show the page to five people who match your customers for five seconds, then ask what the business does and what they would click. Five people find most major clarity problems.
• Session recordings. Watch 20 to 30 recordings of real visits. You will see where people hesitate, where they rage click and where they leave.
• Form analytics. Find the field where people abandon. Removing one confusing field often beats any design change, as shown in our guide to designing a contact form people actually finish.
• Asking customers. Add one question to your enquiry form: what nearly stopped you from getting in touch? Patterns appear within a month.
• Before and after comparisons. Make one change, then compare eight weeks after with the eight weeks before, adjusting for season. It is less rigorous than a split test but far quicker on small sites.

Combine these and you get a clear list of improvements. Then make the changes that several sources agree on, starting with where your main call to action sits, which we cover in call to action placement on a small business website.

Frequently asked questions

How much traffic do you need for split testing?

It depends on your conversion rate and the size of change you want to detect, but most small changes need tens of thousands of visitors per version. As a rough rule, if a page gets fewer than about 5,000 visits a month, only large, bold changes are worth split testing.

How long should a split test run?

At least two to four full weeks so every day of the week is covered more than once, and longer on low traffic sites. Decide the end date in advance and do not stop early because one version looks ahead after a few days.

What should a small business test first?

Start with the elements that change what visitors understand or decide, such as the main headline, the offer, the price presentation or the length of the enquiry form. These produce larger effects than visual tweaks, which makes them possible to measure on smaller sites.