Conversion rate optimization (CRO) is a discipline that has gained huge traction and attention within the digital industry in recent years. When done well, it can help you squeeze more value out of your acquisition channels, find the most effective designs for your site, sell more products and services, and increase the profitability of your business.
But like with any growing trend, among all the priceless advice, you can also find a lot of misconceptions, misleading approaches, and confusion about just how to do CRO right.
So how much do you really know about CRO? Maybe you’ve heard of A/B testing, or perhaps you’re deeply engrossed in full funnel optimization. But beyond the buzzwords, our quiz aims to get you thinking about the core principles and processes that underpin the work of all good optimizers.
Now is your chance to put your knowledge to the test with our quick-fire quiz! And don’t forget, whether you’re top of the class or flunk out on every question, we’ve got plenty of free resources available to help you brush up your essential CRO knowledge.
If you have 1 landing page but want to test 2 headlines, 3 versions of an image, 4 variations of your body copy, and 5 different background colors, how many versions of your page would you need to create?
Answer: D) 120
The math: 2 headlines x 3 images x 4 body copy x 5 backgrounds = 120. The more elements you want to test, and the more variations of each element can quickly grow the number of versions of your page that you would need to display. This helps you learn how different combinations of elements work together, but you can see how the sheer volume of combinations can quickly become overwhelming and unmanageable for setting up manually. Check out this graphic for a guide to how testing all of the common elements that appear above the fold on a typical website can quickly develop into more than a million variations.
What is one of the key advantages of using AI to help with your CRO testing?
1. You no longer need a designer
2. The robots come up with all your ideas for you
3. You don’t need any traffic to get reliable results
4. You can test more variations in a faster time than with traditional split testing
Answer: D) You can test more variations in a faster time than with traditional split testing
AI doesn’t mean robots will be stealing your jobs or doing everything for you. But it does mean technology can take care of a lot of the grunt work at a faster pace than would otherwise be possible. An effective CRO program still needs a solid supply of design ideas backed by strong hypotheses. But AI-powered tools such as Sentient Ascend can help take the hard work out of setting up your tests, with you essentially feeding ideas into it, while the software builds as many combinations as possible, letting the evolutionary algorithms identify the variations with the strongest potential to deliver the biggest wins.
Which of these can have an impact on a customer’s decision to buy from you? Choose all that apply.
1. Product images and videos
2. Descriptions of what you offer
3. Easy to find call to action buttons
4. Testimonials and social proofing
Answer: All of the above.
CRO is no cookie cutter tactic or one-size-fits-all approach that will solve all your website issues, and a key thing to remember is that every site is different. All four answers are typical examples of the sorts of design elements that can have a big impact on a shopper’s mindset and impact whether they decide to buy from you or not. But you need to understand your own customers’ pain-points before you can come up with solutions and optimize their experiences. The research phase is crucial to identify what is most important to your users, while testing allows you to see the impact of addressing these issues.
What is full-funnel optimization?
1. Making sure you accept all major credit cards on your payment page
2. Improving every step of the journey towards your key business goal
3. Getting customers to fill their shopping carts before you ask for money
4. Optimizing your advertisements to get people to your site
Answer: B) Improving every step of the journey towards your key business goal
Instead of focusing on just one page at a time, full-funnel optimization is an approach to CRO that involves identifying a key goal, then working backward from that goal to improve each step on the way. This includes how each step interacts with the next, so can often seem a complex or overwhelming task. Download our white paper for more on how you can get started with full-funnel optimization.
True or false, A/B testing and multivariate testing are the same thing.
Answer: B) False
Both A/B testing and multivariate testing (MVT) are types of split testing, in that they involve testing different versions of a page or online experience against each other to see which performs the best for your users. But where A/B testing involves changing just one element at a time, MVT allows you to test multiple changes to different elements, creating numerous combinations that are tested at the same time. It can help you to achieve optimal results in a shorter time, as you can test more changes simultaneously. It also allows you to explore the impact of different combinations of elements that may have been discounted or missed with A/B testing, which involves removing the ‘losing’ design before going on to run your next test. However, MVT is not always realistic for some sites, as it requires higher levels of traffic in order to get reliable results, and also needs more resource to be able to create and build all the different variations for your tests.
For how long should you run your split tests?
1. 1 week
2. 1 month
3. Until one version has 100 conversions
4. It depends
Answer: D) It depends
How long to run a test for is one of the most common blockers for people getting started in CRO. The notion of statistical significance can be a tricky one to understand, especially if you’re approaching optimization without a statistical background. But in simple terms, statistical significance means running your test for long enough and achieving clear enough results that you can have a degree of confidence that the changes in results are down to the design changes made and not just pure chance. The length of time you need to run a test to achieve a statistical significance can vary depending on a number of factors, such as the level of difference you want to measure, the number of variations being tested, and the amount of traffic you have visiting your site. It is also important to consider seasonal changes and how they might impact your tests, and also leaving your tests to run long enough to cover a full business cycle for your site. This means you get a full view of how the changes affect things on different days of the week and at different times of the day.
If effectively calculating statistical significance is a barrier that is stopping you from making progress with your split tests, then Sentient Ascend helps tackle this by taking care of the calculations for you. It uses a leverages Bayesian statistics within an intelligent learning model, which can adapt quicker than the traditional approaches to statistical significance.