The following are some factors that may affect the conversion rate of a website:
Elements under test - The elements you modify in a test page impacts its conversion rate. Be sure to select elements that are most prominent in the test page. If you modify less significant elements such as footer, you may not notice much improvement in the conversion rate.
Differences between variations – By creating drastically different alternatives for your elements under test (pages, headlines, themes, ...) you may notice huge difference in conversions. If changes in the alternatives are minor (such as increasing the font size, changing some words in a text), the difference in the conversions rate could be minimal.
Number of alternatives - Larger the number of variations, higher the chances of coming up with a winner. However, having larger number of variations required you to run the experiment for a longer duration if your site doesn't have enough traffic.
- Traffic - The higher the traffic in your web is, the quickest you may get significant results. However, you do not need a huge amount of traffic to start testing. Websites with just 1000 monthly visitors can produce good enough results through testing.
Though improvements in conversion rate vary according to type of website, a conversion rate difference of >10% can be considered reliable. In some cases, you may even get more than 200% improvement in the conversion rate.