How much traffic do I need before A/B testing means anything?
Efri · 1 min readIn short
- 1Traffic alone is not the number. Conversions are.
- 2You need at least 100 conversions per variation. 200 to 300 is safer.
- 3So your traffic needed = 100 divided by your conversion rate. If you cannot reach that, do not A/B test yet.
Full answer
The rule of thumb is 100 conversions per variation. Some say 200 to 300 for more confidence. So the traffic you need depends on your conversion rate.
The formula: traffic per variation = 100 / your conversion rate.
Conversion rate
Visitors needed per variation
5%
2,000
2%
5,000
1%
10,000
0.5%
20,000
So if you have low traffic, you need a very high conversion rate to test. If you have low traffic and low conversion rate, you cannot A/B test reliably.
Also, run the test for at least one full week. That covers day-of-week patterns.
If you cannot get 100 conversions per variation, do not A/B test. Instead, make bigger changes. Test one big thing at a time. Or talk to customers to understand what they want. A/B testing is for fine-tuning. It is not for guessing.