Whenever you’re creating web content, it’s easy to make assumptions about what your audience will respond to. Unfortunately, this way of thinking doesn’t always yield optimal results.

A/B testing is a common (and easy) way to test your site’s CRO, if done correctly. If done incorrectly, however, this method can leave you frustrated and broke. So, to save you time and money, here’s what you should avoid when A/B testing your content.

Testing One Element At A Time

While it’s fine to change just one element of your media per test, this approach usually ends up being problematic. For starters, it’s time consuming.

According to Peep Laja, sites that already have a steady stream of traffic take two to four weeks to test. For other websites, it takes even longer. If you lack the resources, testing just one element per month is an unreasonable way to find the trick to a high-converting page.

Small Changes Rarely Bring Big Returns

In general, subtle changes to your page are less likely to yield a big payout. If your business is small (read: does not generate billions of dollars a year) then a 4-5% increase in conversion rate will not make a difference in your return.

It’s All About The Viewers

Your CTA button, copy, and hero might all be in the wrong place, but are you going to test each item individually to see where the problem lies? Hopefully not.

Adjusting a headline or button color is a quick and easy fix, but let’s face it: If your content doesn’t resonate with your audience, then your site will never optimize conversions.

You’re Testing Too Much

While the goal is to optimize your site, it is possible to test too much. If you test two completely different formats with different colors, form fields, headlines, etc., you may find that one clearly works over the other. However, with so many elements changed, you’re likely to not know which was the deciding factor. This is how you realize you’re testing too much.

The solution? Keep the testing variations at a minimum, and maintain a strict record of what’s being altered within each test. At any point, you want to be able to refer to a version and know what made it succeed or fail.

While A/B testing is time consuming, performance marketers all agree it’s worth the effort. Just remember, luck isn’t something that comes along often. If you want to see results, you’ll need to take a methodical, scientific approach that works for your individual business.