Testing two or more variations of one or several pages requires a minimum of visitors to calculate whether the differences on are due to chance or not. If you flip a coin 2 times and get two heads, you obviously can’t say that the coin always shows heads; but if you get 1000 times heads, there is no doubt you should bet head for the next flip! This is exactly the same scenario in A/B testing when it comes to statistical confidence.
If this isn’t a problem for big companies having millions of people visiting their website, it can be a brake to testing for smaller players. Alix de Sagazan, AB Tasty’s cofounder and conversion optimisation specialist, is interviewed by Prestashop, the famous eCommerce solution. She details what websites can do to increase their conversion rates even if they have a limited traffic.