Most advice on A/B testing assumes you have thousands of visitors a day. If you run a small site, a niche blog or a single product page, you probably have a few hundred visitors a week, and the usual “run the test until it is significant” guidance can leave you waiting for months. That does not mean testing is pointless. It means you need to test differently.
This guide explains how to run sensible, low-traffic landing page tests: what to change, how long to wait, and how to avoid fooling yourself with noise.
Why small sites struggle with A/B tests
An A/B test splits visitors between two versions of a page and compares how many take a chosen action, such as joining your list or clicking a buy button. The trouble is that conversion rates are small numbers, and small numbers bounce around a lot. If ten people convert on version A and thirteen on version B, that gap is well within what chance alone can produce.
With low traffic, two things follow. First, small tweaks (a slightly different button colour, a reworded subheading) are almost impossible to measure. Second, the temptation to stop a test early when one version “looks like it is winning” is very strong, and it is the most common way people end up with false conclusions.
Step 1: Pick one page and one goal
Choose the page that already gets the most visitors and has a clear, countable action. A sign-up page with a single form is ideal. Avoid testing pages with several competing goals, because you will not know which one you moved.
Write the goal down in one sentence before you start, for example: “More visitors to this page should join the email list.” Whatever you decide, do not change it halfway through.
Step 2: Test big changes, not small ones
When traffic is limited, you need differences large enough to show up. Good candidates include:
- The main promise. Rewrite the headline so it states a different benefit, not just different wording of the same one.
- The offer. Compare a free checklist with a free mini-course, or a short guide with a template pack.
- Page length. A short page with one form against a longer page that explains more.
- The call to action. Where it sits and what it asks the visitor to do.
Change one big thing at a time where you can. If you rewrite everything at once, a win tells you that the new page is better but not why, which makes the next test harder to plan.
Step 3: Decide the rules before you look at results
This is the step most people skip. Before the test begins, decide:
- How long it will run (at least two full weeks is a sensible minimum, so that weekday and weekend behaviour are both included).
- The minimum number of visitors each version must receive before you look at anything.
- What size of difference would actually change your decision.
Write these down. Peeking at results every day and stopping when the numbers look good is how random fluctuation gets mistaken for a genuine improvement.
A rough guide to sample size
There are free sample size calculators online, and it is worth using one. You enter your current conversion rate and the smallest improvement you care about, and it tells you roughly how many visitors each version needs. For a page converting at a few per cent, detecting a modest improvement often needs thousands of visitors per version. If the calculator says you would need a year of traffic, that is useful information: it tells you to test bolder changes or to use a different method altogether.
Step 4: Keep the test clean
A few practical habits protect the quality of your data:
- Split traffic at the same time. Run both versions together rather than one this week and one next week. Otherwise a newsletter mention or a seasonal change can skew the comparison.
- Keep the traffic source steady. If you send a burst of social traffic to only one version, the result is meaningless.
- Check both versions on every device. A layout fault on mobile will quietly drag one version down. Preview each page on phone, tablet and desktop before launch.
- Do not edit mid-test. Fixing a typo is fine, but changing the layout or offer resets the clock.
Step 5: Read the result honestly
When the test ends, compare the conversion rate of each version and check whether the difference is bigger than you would expect from chance. Most testing tools show a confidence figure, but you can also use a free significance calculator with your four numbers: visitors and conversions for each version.
There are three possible outcomes, and all of them are useful:
- A clear winner. Adopt it, note what you learned, and plan the next test.
- No meaningful difference. This tells you the element you changed does not matter much to your visitors. Move on to a bigger idea.
- A clear loser. You have just avoided rolling out a worse page. That counts as a win.
What to do when traffic is simply too low
If your page gets only a handful of visitors a week, formal testing may not be realistic yet. You can still learn plenty:
- Talk to people. Ask a few subscribers or customers why they signed up, and what nearly stopped them.
- Watch real behaviour. Session recordings and simple scroll or click maps can reveal where visitors lose interest.
- Use before-and-after comparisons carefully. Change the page, then compare a few weeks either side. This is weaker than a split test, so treat it as a hint rather than proof.
- Grow traffic first. Sometimes the best conversion improvement is simply more of the right visitors.
Keep a testing log
A simple spreadsheet is enough. For each test, record the date range, what you changed, your hypothesis, the visitor and conversion numbers, and what you decided. After a few months you will have a record of what your audience responds to, and you will stop repeating ideas that did not work.
Key takeaways
- Small sites can test, but they must test bold changes rather than tiny tweaks.
- Set the duration and sample size before you start, and do not stop early.
- Run both versions at the same time with steady traffic sources.
- A test that finds no difference is still a useful result.
- When traffic is too low, combine customer feedback and behaviour tools with careful before-and-after checks.
Explore more tools and guides
If you are looking for software, scripts, themes and plugins to help build and improve your marketing assets, browse the full range at Altratom, and keep an eye on the blog for more practical guides like this one.
