PLAN YOUR PROJECT
wagento black company logo
PLAN YOUR PROJECT
work-banner.jpg

5 CRO Mistakes and How to Avoid Them

June 25, 2019 by Luis Gomez Morin

5 CRO Mistakes and How to Avoid Them

This entry was posted on June 25, 2019 by Luis Gomez Morin

Conversion Rate Optimization (CRO) is a great way to increase your conversion rates and ultimately increase your company profits. However, the CRO process is not foolproof. If you make mistakes during your CRO testing, this can cost your company both customers and money.

Here are 5 common CRO mistakes and how to avoid them:

1. Tracking is Set Up Incorrectly

When testing a website for CRO, you typically use standard A/B testing. In this model, you post a different version of your website with specific changes, as well as the original version of your website. Find which version has the best conversion rates by setting up tracking on the pages you are going to analyze. Use heat-maps and user tracking, as well as micro-conversion tracking. Micro-conversion tracking tracks clicks and user engagement, allowing you to understand how your customers use your website. This valuable data will help you analyze the success of the different versions of your website with customers.

However, if your tracking is not set up correctly, you will receive incomplete data. You may notice version A has more conversions, but you will miss out on all the valuable micro-conversions that happen on version B. Incomplete data sets may lead you to implement the incorrect website design, and could result in the loss of customers.

2. Not Running the Test Long Enough

One of the biggest downfalls to the CRO process is how time consuming it is. Between research, forming a hypothesis, and creating a new design for your website, being eager to get results right away is understandable. However, for the most informative data, you need to let the test run for an ample amount of time. Unfortunately, there isn’t a time frame that fits all tests.

You should run the test until your data is statistically significant. That will vary based on your customers and when you implement the test. If you have about 1000 customers per week, you will want to wait until data is collected for 1000. Furthermore, if your customers tend to leave items in the cart, but come back days later to purchase, you will want to extend your tests for a few days to match this consumer trend.

Not only should you let your test run for an ample amount of time, but you need to consider when to run it. You should not run it near holidays or slow shopping weeks, because that will not give you accurate data. Be wary of the calendar.

3. Thinking of CRO as Only A/B Testing

Arguably the biggest mistake made with CRO is assuming A/B is the only type of test. While it is certainly the most prominent, there are other tests that can improve your CRO. A/B testing only covers testing a single variable against another to see which performs better. While this test can be helpful, there are instances where it is not useful at all. For example, if you are testing a single page on your website, it could take months for you to garner enough views on that one page for the A/B testing to be successful.

4. Not Understanding the Statistics

You don’t have to be a scientist to ace CRO, but you do have to fully understand your statistics to take advantage of the data you are collecting. One of the biggest mistakes marketers make when first practicing CRO is not letting the tests run their full course. You want your tests to reach their full statistical significance before you pull the plug. It can be tempting to complete studies early to meet a specific time table, but then you are jeopardizing your data.

The difference between 90% statistical significance and 99% is staggering. Think of it like gambling. If you have the chance to be 99% sure you will win the jackpot, why would you settle for 90%? When you stop a test early, there is a 10% chance the statistics you are reading are wrong. That incorrect information could cause you to implement the wrong website variation.

5. Giving Up After a Failed Test

CRO tests are just like any other scientific method - which means you may not like the results. If the test results do not align with your hypothesis, that is okay. Just because the test “failed,” that doesn’t mean it was for naught. Instead, it shows that you have not found the magic key yet. This knowledge can push you in the right direction, so you can hopefully complete a more responsive test the next time around.

CRO tests take up a lot of time. They require a substantial amount of effort and statistical knowledge, but when completed correctly, they are worth the hassle. They can increase your conversion rates by over 200%, which means your business will thrive.

For more information, visit Wagento.com