A/B testing is an effective way to optimize your website for conversions. However, there are some typical mistakes which marketers do during the testing. So we thought we might give a run-down of 30 quick tips which will turn your A/B testing into success!
1. Run A/A tests. Before launching an A/B test, run an A/A test simply to make sure that your testing software works properly. In other words, firstly test the software itself on two absolutely identical page versions. If in the end you receive different results, then study what you may have done wrong or whether the software is reliable.
2. Change the verbiage. Amazingly, even one word may change the conversion rate. Such words as “free”, “trial” are able to increase or decrease the conversion dramatically. In the process you will discover many other words which work the same way.
3. Only facts. Don’t start a test based on your suggestions only. Opinions of millions of users are what should be the primary reason for the test. Otherwise, you will be bound to receive negative results of your A/B testing and only waste your money.
4. Patience. Trust large statistical data. If one of the variations shows significantly higher conversion, don’t stop the test immediately. Wait until the winning variation gets at least 100 conversions (1000 even better). Small numbers tend to change dramatically.
5. Color matters. Sometimes a simple color change may affect the conversion. This is an example of how button color change from green to red increased the conversion by 21%. Of course, it doesn’t mean that you should make all your buttons red. In your case another color may work, so you need to experiment.
6. Some people don’t like reading. Texts are effective on many websites; however, there are cases when video works better (to explain the service essence, for example). If you decide on video, it should be done ideally, otherwise the test will fail.
7. But some people like to read. Be aware of the cases when it is necessary to explain the product in detail, and only after that place “Buy” button.
8. Reputation is important. It’s not likely that users will enter the credit card number on the website they don’t know. Make sure to include the information about payments security, add payment system badges. Having done this, check whether the conversion has increased because you can’t say for sure whether it will work for your website.
9. Macro conversions are better than micro conversions. Even if you have managed to increase the click through rate from product page to shopping cart, you goal is not reached. The only thing which really matters is whether people started to buy more. That is why try to focus on the overall results rather than on micro conversions.
10. Social proof is not always the best option. In most of the cases it is worth to show to your users that many other people also choose your product. Although, there are some examples when social proof decreases the conversions.
11. Don’t be misguided by others’ success stories. If a certain method has worked for one company, it doesn’t mean it will work for another. It is like with buttons – if a red button was efficient on one website, it will not necessary work for other websites. Learn from other success stories, but apply them with caution for your own business.
12. The best results are obtained after the significant changes. Within short period of time little improvements doesn’t make a solid impact on conversions. Sometimes you should not be afraid to be creative and take radical steps.
13. High conversion rate and more revenue are not the same. Surprisingly, sometimes conversion decrease can lead to revenue growth. For example, if you raise the prices by two times, conversion will go down by 20%. However, you will still earn by 60% more. Optimization is worth to tackle to increase the revenue, not conversions.
14. Mondays are not Tuesdays. Your website visitors on Monday may differ from those who come on Tuesday or Saturday. So you should measure the effectiveness not on the daily basis but by comparing the whole weeks. Users’ behavior may essentially vary during the week and distort the results.
15. Don’t be afraid of bounce rate. Many marketers are afraid of this metric. In fact, it is not so important what it shows as long as your conversions grow. Sometimes ignore secondary metrics such as bounce rate, because they do not affect the sales directly.
16. Separate the tests. During the A/B test planning make sure to prepare separate versions for mobile and desktop computers. Mobile user experience is different and may influence test results.
17. Less doesn’t mean better. Decreasing the number of steps of conversion path might not lead to the conversions growth. As the matter of fact, many case studies show that more stages increase the conversion rate.
18. Don’t make people think. It is rather a misconception to think that your website visitors know what they want and what they should do. Let the people just press the buttons instead of selecting the option in the drop down menus. President Obama’s administration used this method to increase the conversion of fundraising goal by 5%.
19. Customize the interface. The best buttons with call to action are not those with generic “free trial” or “order now”. Call to action should correspond to the product you are selling.
20. Images have got the power. Pages with images convert by multiple times better. Moreover, some pages are able to generate higher conversion than others. Use the tests to know what images yield better results.
21. Navigation bars affect the conversion. It is not always the best option to let the users use the navigation. This example shows that when navigation bar has been removed from landing page, conversion rate has grown – users simply had no other chance than click “Buy”.
22. Simplicity beats persuasion. If after landing at the page the user clearly understands what they are expected to do the next it is a key to success. It is even better than the page with excellent selling text. Ideally, these two factors should be combined. Please note, that when you are writing a copy you need to make sure that you do not mislead the user, because then you will receive the negative outcome.
23. Focus, focus and focus. Each page should have one purpose only. Pages with the single powerful idea convert better than pages which are aimed to solve several tasks at once. For this reason it is advisable to create landing pages for each purpose separately.
24. Testing doesn’t end with your website. Apply the A/B test principles all the way: test ad design, Twitter profile image and change in number of followers. Everything can be tested.
25. Ugly design can convert. Website with poor design does not necessary give low conversion. Look at these websites, especially those which make millions dollars.
26. A/B testing doesn’t affect SEO. Many marketers are afraid that A/B test may have a negative impact on search engine optimization because of the page content change. Make sure to follow Google recommendations and everything will be ok.
27. Decrease the friction. Don’t make the user do things they should not do. For example, fill in unnecessary fields, confirm credit card address, etc. Test the website version with the minimum number of fields and see how conversion will change.
The last advice we would like to give you – don’t trust anything except figures when running an A/B test.