Conversion rate optimisation experts are the best people to ensure you’re doing everything correctly. They usually rely on tools such as Optimizely or Visual Website Optimizer (VWO), which are recommended for this type of work, but others are out there, too. But before that, another critical step is ensuring the design you want to test is high quality and in line with your brand standards. 

Step Four: Implement Design

The next step is to design your testing. There are lots of conversion rate optimisation experts out there who will help you to conduct split tests by developing and amending a copy of your current website and offering different domains to each visitor during their search. The way you do this will depend on where the majority of your traffic comes from and how complex your website is. Don’t forget that you might want to test each website on the different browsers to see if the uptake is different. Different demographics do tend to gravitate to certain browsers. 

Using a consultancy service will allow someone to walk you through your testing, even if your website is more complex to ensure the design and links all work correctly. They can also assist in directing a certain percentage of your traffic to specific test pages depending on what you’re wanting to test. Ultimately, you can test as many pages as you like, but bear in mind that the more you test at the same time, the less accurate your results are likely to be, as you can’t be sure which test the results relate to. 

If you already have a fairly high traffic amount at the moment, but just need to focus on the conversion of that traffic, it’s best just to send a small portion of your current traffic to your test design. Stick to 10%-20%. You don’t want to lose your traffic flow if the test group don’t like the new design. 

If you choose to use Optimizely or VWO you can also link your test site to Google Analytics. This allows you to see the average order price for both your standard and test websites and also allows you to see the split of demographics by conversion. From this, you’ll see what types of people prefer your test version and who might prefer your older version. 

Don’t be disheartened if your test site doesn’t have much effect or if you find your conversion actually decreases in comparison to your usual website. You can always try again. There is an unlimited number of tests that you can do, so just keep at it, you’ll find one that works eventually. 

Conclusion

You might even find that your conversion decreases, but your order value actually increases, indicating that the types of customers who are wanting to spend more money prefer this test version. Don’t overlook wins like this; they could be crucial to you understanding your customer base. You may even want to change the angle of your website so that it works to convert more of this type of customer. 

Within the implementing phase you should have:

• Obtained sign off for testing from your client or investor (if required)
• Started to test the variations of your website against your hypothesis
• Tested your website on multiple browsers
• Tested different versions of your site by demographic
• Review