Know about the Ways to Avoid Some Common Mistakes in Your CRO Projects

For those who work in CRO (Conversion Rate Optimization) and UX testing projects, it is a common issue to get confused due to excess information. This information is related to metrics, reports, implementation, testing platforms, and formulation of hypotheses.

CRO mistakes that you should avoid

1. Improper testing schedule: Timing of conducting the test is important, as also the timeline for various phases like designing,launch, results, etc. Ideally, the tests for SaaS should not be conducted during holidays or vacation seasons when most of the target group would not access the websites. You may get results that are not reliable.

2. Lack of information about the project: You need to have a clear understanding of the objective, goals, KPIs (Key Point Indicators), and
the project’s desired results to make the project successful.

3. Ignoring micro-details
These details range from observing form-fillings, clicks, etc. For example – if you are working with an e-commerce website and would like to analyze customer behavior, you can track details like:

a) products added to cart and discarded
b) number of visitors viewing the product pages
c) visitors who surf through all the pages of your website
d) individuals who have created accounts

4. Inadequate sample size
A/B testing or split testing deals with comparing different variants of website visits to understand what leads to more conversions.Now, this A/B testing would not give you accurate results if it is conducted on a page or website which attracts a low volume of visitors. This testing requires large data. So, you can wait for the page to receive the desired high volume of visits. Till then, it would be better to put the test on hold.

5. Faulty interpretation of test reports
A proper understanding of the test reports is important to check whether all the goals have been achieved. A decision can be taken if all the desired results are obtained. Overemphasis on a particular aspect and ignoring others is not proper.  The best method would be to segment the entire target population into different groups, study the details and results of different segments of the target group and then arrive at a decision.

6. Ignore good test results and giving up after a failed test
If you find the minimum or no negative points for a particular webpage after a test result has been conducted, it does not mean that you should stop researching. There may be several areas that have not been looked into. Also, you should conduct CRO tests as a continuous process for developing or improving the webpage.  In case of a failed test, you should not give up but try to find out the reasons for failure and overcome this in all following tests. There can be several reasons like a wrong hypothesis, improper segmentation, etc.

7. CRO is not A/B testing
A/B testing is one among many other CRO tools. A/B testing and CRO are not the same. Under A/B testing, you can compare one variable against another based on the superiority of performance.

8. Unorganized one-off testing 
It would be good if you, as a marketeer, analyze your website’s performance before conducting A/B testing. An organized process is bound to give better results. An organized process consists of a roadmap, KPI benchmarks, hypotheses, and analyzing tools.

9. You’re not running the test for long enough
You have to run the test at least up to the time when it reaches statistical significance. You can also extend the testing beyond this point. This is important because you may get some conversions from visitors several days after their initial visit to the website. There can also be traffic fluctuations during various periods.

10. Not tracking all types of conversions separately
Conversions from different channels like desktop or mobile should be tracked separately. This is because if you run a test on the results obtained from these, the outcomes may differ. It would not be proper to consider the combined results in these cases, but the results obtained separately from different devices.

Conclusion:
While conducting a test, start systematically. Schedule the processes with care and check for minute details. If you have already conducted tests earlier, try to take forward the learnings and implement them in future tests.
CRO is a complex process, and an organized method would save your resources and minimize mistakes.

• March 18 , 2021
• Rushik Shah