If your website isn’t the #1 salesperson within your organization then you’re simply leaving money on the table. Every savvy marketer knows that clicks and conversions count in the world of online marketing. You should know how many people visit your site everyday, what pages they’re looking at most often, where they enter, where they leave, and a host of additional metrics that give you insights into the behavior of potential customers. Once you have this knowledge, you can use it to maximize your lead generation and to get a tremendous return on your initial investment. Ready to turn your website into the greatest salesperson of all time? Then get to know split testing.
Split testing is a procedure performed by marketers and researchers to discover what’s most popular among test subjects when those subjects are choosing between 2 or more similar options. In most cases, the differences between the options are minimal. These slight differences make it easy to see what variables(s) people prefer. For instance, let’s say we want to design the perfect billboard ad for our product, a Porsche GT3. To do this, 100 viewers are shown two pictures, each of a Porsche GT3. Porsche 1 is green and Porsche 2 is red. If we ask these viewers to choose which Porsche they prefer and 75 out of 100 viewers choose Porsche 2, then we know that our test subjects prefer Porsche’s that are red rather than green. The key variable is color. We can use this information to test other elements such as shape or text for the ad. If we were to conduct another test, we would make both options red and change a different variable within the images. By following this process over and over, we will eventually come up with the best possible ad for our billboard, based on what our test subjects prefer.
A split testing approach works excellently when trying to maximize your website’s performance and increase website conversion rate. The goal is to design web pages in the most optimal manner possible so that the highest number of visitors take a desired action. If for instance you want to get more people to sign-up to your company newsletter, it’s prudent to split test the sign-up page as well as other pages that lead up to the sign-up page.
When split testing a web page, you want to test the following elements:
- Page layout
- Headline
- Subheadline
- Call to action
- Quantity of text
- Copywriting
- Images
By testing all of these variables, you’ll eventually create a very highly desirable web page for your viewers and more importantly, increase website conversion rate. To test, you’ll need to start by choosing one (1) of the elements above and making it the test variable. Let’s say you choose to test the headline of your newsletter sign-up page. You will need to have two identical sign-up pages in which the only difference is the words written in the headline. Once both pages are created, you would simply send an equal amount of traffic to both pages to determine which page converts better. In this instance, a conversion would be someone signing up to your newsletter. Assuming you have a sufficient amount of traffic to both pages, you will be able to clearly which page converted higher. From there, you would then repeat the process for the other elements, using the higher-converting headline in all future tests.
To put split testing into action, you’re going to need a split testing tool. There are many tools out there, but I personally prefer to use the Experiments feature in Google Analytics. This tool is great because it’s free, it integrates seemlessly with all of your other Google Analytics data, there’s great reporting, and you can test multiple variables at once. The only draw back here is that you have to have access to the internal workings of your website in order to set up the test and use the tool. Therefore, you’ll either need to know how to create new web pages and how to modify html code or you’ll have to consult with a web developer.
Your other option is to use Visual Website Optimizer. Now this is a great tool because you don’t need any technical coding skills whatsoever! The VWO team has made the entire process GUI-based so you can easily change any of the elements I mentioned above simply by dragging and dropping. It also comes with a suite of additional features, such as heatmaps and behavioral targeting, that are really helpful. The only drawback with this split testing tool is that there is a monthly cost. That cost starts at $49/month on the low end and can be more than $250/month depending on how much traffic you’re getting.
The last thing to keep in mind when using a split testing tool is that split testing is only beneficial when you have an ample amount of data. In other words, if you’re testing the effectiveness of an element on your web page, you’re going to need a respectable amount of traffic before you can come to a rational conclusion on which version of the page should be used. 10 visitors in a test is not enough. 100 visitors is about as low as I’d go. That will give you 50 unique views to Page A and 50 views to Page B.
That’s all folks! Get out there and start testing. It’s a great way to increase website conversion rate and it can make a world of difference to your company’s bottom line.
