Convert More of Your Traffic with Multivariate Testing


If you work in an online industry, you’re probably familiar with the terms subscribers, bounce rate, conversions, signups, and sales. Getting your website to perform for you and reach your goals is no easy feat. Luckily, tools are available to take guesswork out of the equation when it comes to what works on a website or email campaign.

Multivariate testing can be performed to determine what elements have the biggest impact on conversion. This tool is essential in getting the subscribers, sales, and signups you’re looking for. In this article we’ll take a closer look at how multivariate testing can work for your website.

Online markets where multivariate testing can work for you include home loans, SEO, insurance, and a plethora of very competitive industries. Multivariate testing ensures you get the most out of your marketing dollars and are heard above the competition. Some simple tests could essentially double your revenue as the result of doubling your conversions.

When to Conduct Multivariate Tests

The right time to conduct a multivariate test is when you have an online page or campaign with a clear and measurable goal. Perhaps you’re trying to get people to sign up for your newsletter on a landing page, or want to increase sales on your e-commerce site.

Once you’ve identified this quantifiable goal, running these tests will give you a better idea of how fast conversion improvement can be made in the shortest period of time. Another great time to conduct tests is when you want to implement a new feature – you’ll learn the impact it has on visitors.

Factors that Influence Reliability

Realize your first multivariate test isn’t likely to give you substantial results. The key is to test regularly, keeping the following in mind:

Number of Visitors:

Reliable results can be attributed to a higher amount of visitors. To estimate how many visitors will be required for your test, use the split and multivariate test calculators available online.

Conversion Rate:

In general, results for pages with a low conversion rate (say 1-2%) take a much longer period to produce significant results than pages with a higher conversion rate (say 40-50%).

Difference in Performance:

When the difference in performance is large (say > 10%) you’re more likely to see accurate results than when that difference is small (0.5% or so).

Tips for Conducting Multivariate Tests

• Set dates when you can review the reports of your tests. Track the reports on a spreadsheet to ensure your testing the right elements each month.
• Test the same elements on your site or campaign as the competitors in your niche.
• It takes time to get the results you’re looking for. Implement multivariate tests on high-traffic pages of your site and determine how long you want the test to run. The length of the test is usually determined by the amount of traffic your website is receiving.


No matter how skilled a designer you are, nobody can come up with a perfect site or email campaign with maximum conversions on the first attempt. This is where multivariate testing can achieve the success your website is looking for through conducting simple tests on the elements of your page. If you’ve used multivariate testing in your own site or campaigns, tell us your experiences in the comments below