How to get your start in conversion optimization (Review)

Last week marked my first week of the CXL Institute’s Conversion Optimization mini-degree. The course is divided into 5 tracks. The first track is the foundation which essentially covers diverse topics such as copywriting, psychology, design and google analytics, setting a good base for future tracks.

The track starts with an overview by the conversion scientist himself, Brian Massey. If you don’t know who Brian Massey is, you should check out his podcast “Intended Consequences” where he covers topics related to conversion optimization. In the lessons, he lays a lot of emphasis on data driven hypotheses building.

Best practices should only be used as a starting point if you really don’t have deeper level website and customer data. While extensive data analysis doesn’t guarantee winners, it improves your ability to test what is likely to work, thus improving the overall win rate of your optimization program.

Here is how he explains one should go about gathering data to back-up hypotheses:

• Has anyone else tested or made the same change and seen a lift in conversion rate (or other metrics)? While that may not be applicable to your business, it is still better than no data at all.
• Is there any existing research conducted by your organization — surveys, marketing research studies, personas etc. — that can be referenced? This is a good time to review the big and small studies undertaken.
• What does your web analytics data indicate? Does anything pop out when you analyze the actions users take — next page navigation, bounces, device, channel etc?
• Do you have tools like hotjar and mouseflow implemented on your website that provide heatmaps, scrollmaps and screen recordings? These tools provide a lot of context and color around the data you see in your web analytics tool. For example, your web analytics data may just tell you that the bounce rate on your homepage is high but not where the problem could potentially lie. Using heat maps and scroll maps, you can see how users are engaging with the page, how much they are scrolling and where they are clicking/ mousing over, thus making more informed hypotheses. So, if users do not scroll down a page (i.e. the page has low scroll depth), important next steps will be to test and optimize the value proposition above the fold.
• Have you conducted any user testing, the findings of which corroborate the hypothesis? Tools such as usertesting.com and Validately help you test users remotely. You can choose to moderate the session and interview users as they go through your website, app or prototype. Or go for the non-moderated option where users record themselves (and their thoughts) while completing the tasks you have set-up.
• While we are mostly focused on getting exit intent insights from users that are abandoning checkout or the website as such, Massey suggests asking users who successfully complete a transaction “what almost kept them from buying?” This is powerful as these users are still in the moment and willing to possibly share a little more of their time. They have also had a positive experience (likely) with the brand and may want to help.

There is also a lesson dedicated to the fold and ideal page length which I found very helpful. Although looking at social media feeds on our mobile devices has got us habituated to scrolling, the fold isn’t dead. The “fold” is the top portion of a web page that people see without having to scroll down. And hence, above the fold should have the most critical stuff you want users to know about and the most important actions you want them to take (ideally).

It does get tricky to accommodate CTAs above the fold in forms due to the number of fields and the field length. However, there are other factors that can contribute to a form’s success much more than placing the CTA above the fold. Optimizing the number of fields, form layout, making sure the information sought is as per user expectation, providing helpful instructions or tooltips, displaying inline error validation, can go a long way in making the form easier to complete.

There’s no universal page length that marks the fold so look at the share of the top resolutions by device in your web analytics tool and make sure the content above the fold is optimized for those. Ideal page length also differs based on context. For transactional sites, ideal page length is the amount of content people need to take action. However, a big benefit of shorter pages is faster load speed.

The CXL program is dense with content and the next few weeks will certainly mark a pretty steep learning curve for me. Additionally, I found some useful resources on the TLC slack channel in a thread about “a starter pack for those interested in or new to A/B testing and optimization.” BTW, if you are somewhat serious about testing & optimization, you should join the Test & Learn Community (TLC) run by Kelly Wortham. Some of the recommendations were:

Education & Skills

• Strong understanding of web development technologies (HTML, CSS, JS) through freecodecamp.com, in particular understanding state management
• CXL CRO minidegree
• Introduction to statistics
• Certification in one or more testing tools

Experience

• Start testing within your existing toolset. Google Optimize is free to use. Facebook allows testing too.
• Start understanding user behavior on the website by way of tools like HotJar
• Gain experience on more than 1 testing platform
• If this is currently not possible at work, then look for pro bono engagements for non-profit and small businesses.

Communities

• TLC on Slack
• Measure Slack
• Digital Analytics Association
• CXL Facebook group

Resources

• https://www.alexbirkett.com/ab-testing/
• http://www.experimentationhub.com/a-b-sensei.html
• http://exp-platform.com/
• https://www.analyticshour.io/
• https://craptalks.com/
• Conferences: https://effectiveexperiments.com/blog/cro-conferences-to-attend-2020/

Books

• Trustworthy Online Controlled Experiments: A Practical Guide to A/B Testing by Ron Kohavi
• List of books recommended by Alex Birkett

My week ended with the first ever Virtual Digital Analytics Hub, where it was really interesting to discuss real world strategies for optimizing your optimization program and driving adoption across the organizations with stalwarts in the space. More on those discussions in my next post!

Stay tuned.