How To Determine Your Next CRO Test – Part 2

By , CRO Manager


Last month, we discussed the struggles with determining your next conversion rate optimization (CRO) test. If you missed it, see Part 1 where we covered Google Analytics, AdWords, and heat mapping.


This month, we’ll take a look at a few more factors that will help you determine your next test, including:


  • Video recordings
  • Usability testing
  • Multiple tests & what to test first
  • E-commerce vs. lead generation


Video Recordings & Usability Testing


UserTesting can be used to create videos of participants using your website. The key to these videos is not to ask specific questions but to observe users as they navigate the site. You should ask these participants to complete a task, such as purchasing an item or completing a lead generation form and observe as they go through this process. These videos will show you where users are experiencing frustrations or functionality issues.


Once you’re aware of the areas of your site causing problems, you can focus your testing directly on these items. You can use these videos as building blocks from what you have learned in analytics and from your heat-maps to determine your next test.


In the following example, we saw this funnel in analytics. We noticed that we were losing 10%+ customers after the address step of the checkout process. By creating UserTesting videos, we found multiple users stumble at this stage in the process, seeing the error message on the bottom. By using these factors as building blocks, we are able to make the assumption that this friction experienced by the users is the cause in the 10% drop in the funnel.






In addition to videos, you can also utilize usability testing software such as Usability Hub. With usability testing, you can perform tests such as:


  • Five-second tests
  • Click tests
  • Navigation tests
  • Preference tests


These tests will help you see how accessible and usable your site is. You can ask a specific number of participants about certain pages of your site.


Five-second tests are going to be the most important of these tests. Users form their initial impression of a page within the first 5 seconds so it’s important that they know what your site is about within these 5 seconds. If they don’t find what they’re looking for, they’ll bounce. Therefore, select a number of participants to look at your page for 5 seconds and see if they can determine the purpose of the site.


Click tests and navigation tests will tell you how easy it is for users to navigate through the flow of your site. You will be able to analyze where users are clicking and if they successfully navigate through. By looking at a certain sample size, you’ll be able to see if there are commonalities between participants and if there’s a certain step users get lost at. This will help you determine if there’s a step in your process that needs to be clearer.


Another type of test looks at participant’s preferences. This will help give us data behind how your potential tests will perform. You can provide participants with multiple screenshots to determine which one might perform better.


What To Test First


Last month we mentioned that you might have multiple tests to choose from. How do you choose which one to test first? Here’s where you need to consider your analysis. Remember everything you found in Analytics, AdWords, heat maps, UserTesting videos and usability testing. Where are users experiencing the most friction? Are there functionality issues that need to be addressed immediately?


Make the changes necessary to reduce functionality issues. Don’t worry about testing these. If users are unable to make a purchase or navigate your site because something is broken or not functional, there is no need to test. Make these changes and then take a look at the remaining testing ideas you’ve generated. Which test will have the most impact? Begin with the test that is most likely to have the highest impact.


Once you begin running tests, you’ll always learn new insights from your test results. Whether your test won or lost, analyze your findings to generate new insights. These could lead to future tests or ways to alter the test you just ran in order to further optimize your landing pages.


To determine which test will have the most impact, not only should you consider your users and the analysis performed, but you also need to consider what type of client you’re working with. This leads us to our next point, the differences in testing ecommerce versus lead generation.


Ecommerce vs. Lead Generation


When considering conversion rate optimization, you want to start by focusing on the conversion funnel. When you’re performing your analysis and using these different factors and platforms to analyze user behavior, you want to start by making sure the conversion is easy to complete.


If you’re an ecommerce client, you want to ensure your shopping cart is easy to use. Ask questions such as:


  • Does it cause friction with your users?
  • Do your users trust your site and the checkout process?
  • How long is the checkout process?
  • Is it clear what payment methods are accepted early in the process?


These are important aspects to consider when determining your next test. If they’ve gotten all the way to the shopping cart, you don’t want to lose that conversion. If we know from our analysis that there aren’t any huge issues prior to the shopping cart that are losing customers, then let’s focus our efforts on the cart to ensure we capture those conversions.


If you’re a lead generation client, I bet you guessed it. You should focus primarily on the form your users need to fill out. Again, prior to this focus you need to look at your analysis and see if there are friction areas elsewhere that are causing people to leave your page. Once we know the issue is the form, here is where we want to optimize. Is every field necessary or can we shorten the form? Is it obvious to users which fields are mandatory? Is the call to action compelling and easily accessible? These are all things you want to consider when optimizing for leads.




I hope it’s clear how each of these tools is a building block in determining your next CRO test. By utilizing these tools and analyzing your landing pages, you should be able to plan a pretty detailed roadmap for the next few months of testing. Use what you learned to determine what will have the most impact and begin with those tests. As you begin running tests, you’re sure to generate new insights and new testing recommendations.


If there are other tools or factors you’re utilizing to help determine your next CRO test, please feel free to share in the comments section!

Twitter Facebook LinkedIn Google+ Email Print More

3 thoughts on “How To Determine Your Next CRO Test – Part 2

  1. Greg Ahern

    Hi Samantha,

    One tool I use often are simple surveys and feedback tools. Not only do you learn what users are getting stuck on and errors on the site but you can gain insights on what customers are looking for and desire. This helps align the content and navigation to fit what the user wants. It can also provide information for new products and services. Both surveys and feedback / need help tools can have an active or passive engagement with the user depending on the site. Lastly they can provide information on specific device and browser related usability issues.

    Disclaimer – I am the founder of which provides a suite of CRO and lead generation tools.

    1. Samantha KerrSamantha Kerr

      Thanks for reading, Greg! I’ll have to look into Ometrics. We do currently utilize HotJar for feedback. We don’t use it often but we have occasionally used the feedback polls which we have given us pretty good insights into what the customers are looking for and problems they may be running into.

      Do you utilize other tools / platforms for your surveys or feedback tools or just Ometrics?

      1. Greg Ahern

        I built Ometrics because I was tired of using so many tools not only from the cost standpoint but just having to access each account, so I combined the tools into one service in 2013.

        With surveys you can create a poll using radio buttons or the emotion icons. The data is presented for each submission as well as total results like 50% selected “Yes”. We recommend always having the feedback or need help tab up when not surveying. If you want to get answers to a particular question then the popup slide survey has a higher conversion rate since it is active not a passive tab. The popup can be triggers by different user actions such as exit, second time to site, time on page etc. These same tools can be used to provide offers or coupons.

        I personally find the one question survey is most effective, although Ometrics can do multi question surveys. So at this point I do not use any other tools for surveys and feedback it does what I need on desktop and mobile… although I have a list of “nice to have” features I want to add :-).


Leave a Reply

Your email address will not be published. Required fields are marked *