My Learning Journey of Conversion Rate Optimization
There is an interesting case study Brian Massey, Conversion Scientist at Conversion Sciences, shared about a website re-design project done by FinishLine.
On November 19, 2012, FinishLine updated its website from A to B. With more stylistic images, minimalistic category pages, high-quality photography, and celebrity endorsement. What is your instinctive prediction of this launch?
Well, the reality is, within just a few weeks, the business and this website had lost $3 million. That could be any marketer or business executive's worst nightmare.
As a marketer, I'm not new to the shocking feeling of surprise negative results of a new campaign that is supposed to bring success and excitement to the team. And I don't think I'm alone. The question is, how does it happen? And how to avoid it?
Brian suggests that "the culprit" of wrong decisions is our biases. For example, confirmation bias makes you only pay attention to things you agree with or think are good. Novelty bias grabs your attention to believe the cool things are the best things.
We clearly don't want these biases to influence our judgment. That's why we should rely on data and science. We should look into data to understand what our visitors, customers, and prospects want instead of what we think is good.
I believe that is the mindset any marketer should have if one wants to increase the conversion rate of any marketing campaign. And there are just so many things to touch on. So I decided to take on a learning journey with CXL, and I'm starting this "journal" to document things I learned. Feel free to comment and let me know what you think.
4 takeaways to lay the foundation of Conversion Rate Optimization （CRO)
Choose wisely what to testa and test with good sample size and quality. Figuring out what you don't test is just as important as figuring out what to test because your time and traffic are limited. If you're doing a quantitative test, you should make sure to increase the sample size that can yield meaningful insights. If you're doing a qualitative test, such as bringing a focus group, finding people trying to solve problems that our products and services solve is much more effective than finding a random audience.
Create an idea list and rank them based on impact, confidence, and effort to test. We got inspiration and test ideas from colleagues and managers all the time. No idea is a bad idea, but some ideas just won't make your priority. How do you manage these ideas? Try creating your idea list. And write a little hypothesis to specify the idea with a formula like" If I ... I expect ... to happen, as measured by ..." Then rank and assign points to your idea based on its result impact, confidence or proof, and effort to make it happen. Those with low effort, high impact, and high confidence scores should jump on your priority. If you're curious about how the Framework could look like, here is an example.
Identify your sources of insight. Just like journalists write an article, the sources of information determine if the article is credible and serves the truth. When we gather insights to support a marketing campaign idea, here are some sources we can look into. First, Google. Because someone else might have done a similar campaign or test. Second, refer to your idea list and identify what the primary concerns are. Third, always think from the perspectives of your personas. Last but not least, use analytics tools, like Google Analytics, Heatmap, eye tracking, to gather user intelligence.
Have a certain tolerance to failures. Most tests are a failure. In fact, many times you go through are going to not deliver the "ah ha" moments. And you and your team need to be ok with it so you can get a momentum of experimentation. As you get better at researching your ideas, you'll do a better job not testing things you shouldn't be testing.
Most valuable best practices for a B2B marketer
There is a long list of things that can be optimized, but I found these most valuable best practices for a B2B marketer to deal with daily.
Web forms
When we look at forms, the top goal is to reduce friction. Here are some ways to do it:
Set clear expectations. Most people may like surprises in real life, but not when they try to fill out a form. If you're transparent about what's going to happen, there is less chance the form-filler will be (negatively) surprised when it happens. A couple of examples:
"Take only 20 seconds" -> expectations for how long it takes to fill out a form.
"Fill out this form and get instant access to X" -> expectation setting for what comes next.
An Eloqua study found the "sweet spot" for form length is between 5-10 fields, where 7 was the optimal number.
You can intentionally increase friction to improve lead quality and get better lead scoring.
If you really do have a long form, break it into 2 or 3 steps. Autodesk tested a two-step form, which broke a long form into two parts (while keeping all the form fields), increasing the free trial sign-up rate by 23% at a 99.9% confidence rate.
Pre-select what you can and try auto-populating information if it allows.
Buttons and CTAs
Visitors take action depends on (1) if they notice the CTA; (2) if the next step is obvious; and (3) if they see value in the next step. You will be surprised how invisible the main CTA buttons are on many sites. You should have one primary CTA on your landing pages, and a button is usually better than a link. When you write a CTA, be sure to include a trigger word, and here is a blog post by Ott Niggulis on the details of mastering the CTA.
Persuasive design
You probably think design is the realm of designers. Coming from a freelance designer background, I think it's beneficial for a marketer to understand persuasive design because human beings are visual animals. People trust what they see more than what they hear. Peep Laja, Founder of CXL, did a presentation for SearchLove Londer and introduced five principles of persuasive design.
Clarify above all. Your page should be able to answer these questions when a new visitor land on your page: What is it? What can I do here? How is it useful to me? If you can't deliver clarity in a second or two, you lose your audience.
Visual appeal. We take 50 milliseconds to make a snap judgment about a product, a person, or a website. And we all know that if the first impression goes wrong, it takes so many more interactions to change one's mind.
Strong visual hierarchy. It's quite simple. We make what's important bigger and stand out with color. Try to rank the elements on this webpage. Then go look at your own site. Is there something important that visitors are likely seeking but isn't high enough in the hierarchy? Try to get into this habit of scanning other websites. That's my new leisure game during the breaks.
Thank you for reading and making it this far. I will be sharing more learnings and thoughts around writing value props and marketing copy in the next post. And I'd love to hear from you about any feedback or comments.
Till next time :)
Zoey
Comments