A/B testing is widely considered the best way to improve conversion rates. By sending visitors to two versions of a page, you can see which one works better and use real data to guide your decisions.
However, A/B testing only works if enough people see each version to get reliable results. If you test with too little traffic, you might make decisions based on chance instead of real improvements. In this guide, we’ll explain why traffic volume matters, how to find your own threshold, and what to do if your site isn’t ready for testing yet.
Why Traffic Volume Matters for A/B Testing
A/B tests compare two conversion rates to see whether the difference is real. If your sample size is small, random changes can make results look important when they’re not. Statistical significance shows how likely your results are real. Most testing tools require 95% confidence to pick a winner, which means you need enough visitors and conversions.
The risk of calling results too early
One common mistake is checking results every day and stopping the test as soon as one version seems to be winning. Early leads often disappear as more data comes in, so a version that looks like a winner at first can end up losing. Decide on your sample size before you start and stick to it to avoid this problem.
How to Work Out Your Traffic Threshold
No single traffic number works for every website. Your threshold depends on multiple factors. Knowing these helps you decide whether running a test makes sense before you invest time.
Your current conversion rate
Pages with higher conversion rates need fewer visitors, since each visitor is more likely to convert. For example, a page with a 5% conversion rate will reach reliable results faster than one with a 1% rate. Always check the conversion rate for the exact page you want to test, because your site-wide average might not tell the full story.
The size of the uplift you want to detect
It’s much harder to prove small improvements than big ones, and you need a lot more traffic to do it. For example, finding a 10% improvement takes almost four times as many visitors as finding a 20% improvement. That’s why sites with less traffic should focus on testing bigger changes, like layout or messaging, instead of small tweaks like button colours.
The confidence and power you need
Confidence tells you how unlikely your results are to have happened by chance, while statistical power shows how likely you are to find a real effect. The standard is 95% confidence and 80% power. If you lower these, your results become less reliable. Most sample size calculators use these settings, so they’re a good place to start.
What the numbers look like in practice
The table below shows how much traffic you need for a basic two-version test at 95% confidence and 80% power. The last column assumes 5,000 visitors a month to the page you’re testing.
| 2% | 20% | 21,100 | 42,200 | 8.4 |
| 3% | 10% | 53,200 | 106,400 | 21.3 |
| 3% | 20% | 13,900 | 27,800 | 5.6 |
| 3% | 30% | 6,500 | 12,900 | 2.6 |
| 5% | 20% | 8,200 | 16,300 | 3.3 |
As a general rule, we recommend at least 5,000 monthly visits to the pages you want to test before starting A/B testing. Even then, focus on changes that could increase conversions by 20% or more. If a test would take longer than eight weeks, reconsider, since seasonality and cookie expiry can affect your data. No matter your traffic, run each test for at least two full weeks to include both weekday and weekend behaviour.
What to Do If Your Site Isn’t Ready for Testing
If your site doesn’t meet the traffic threshold, you can still improve your conversion rates. Sites with less traffic should rely more on research and less on statistical testing.
Fix the obvious problems first.
User research can reveal problems that don’t need testing to confirm, like broken forms or confusing navigation. Watching session recordings or running usability sessions with just five or six people can quickly show you the main issues. Fixing these right away is often faster and more valuable than running tests.
Measure before and after
If you make a change without running a formal test, compare your results from similar periods before and after the change. This approach isn’t as precise as A/B testing, so it’s best for bigger changes, and you should consider seasonality. Track both lead quality and quantity to see whether your change is making a difference.
Test higher up the funnel.
If you don’t have enough final conversions to test, measure a more frequent step, like clicks on your main call to action. These micro-conversions help you gather data faster, but they should still be clearly connected to leads or sales. Use them as an early sign of progress, not as the final result.
Grow the traffic you already have
Sometimes, the best way to start testing is to bring more qualified visitors to your important pages. Increasing organic traffic to your main landing pages with SEO helps you reach the testing threshold while building your sales pipeline. That’s why we plan SEO and CRO together, instead of treating them as separate projects.
Choose the Right Approach for Your Traffic
A/B testing is a great tool, but it only works if you have enough traffic for reliable results. Figuring out your threshold before you begin saves time and helps you avoid acting on bad data. If your site isn’t ready yet, research and careful measurement can still drive big improvements.
Our CRO team can look at your traffic and suggest the best approach for your website. Contact us to find out if your site is ready for A/B testing and where you should focus first.
