Understand user behavior and use an evidence based model for experimentation: Ruben De Boer
Want to be on GrowthFit?
We feature D2C operators and CRO experts with a story worth sharing.
We feature D2C operators and CRO experts with a story worth sharing.
My journey started ten years ago, though it wasn’t my job title yet. I had set up my experiment over ten years ago. I started with conversion Rate optimization because that combined several of my passions- human behavior, data analytics, goal oriented and result driven. That’s what really interested me then. I made all the possible mistakes initially in my career. But gradually, I started growing and learning more and more. My interest is in human behavior and products. I get to experience the culture of experimentation with various organizations.
I also believe that if you find your passion in your work, you should also use it to inspire and help others. That’s why I do online coaching and create courses.
I am lead conversion manager at Online Dialogue with a multi disciplinary team. It was founded in 2009. We do only CRO and enable digital transformation for our clients.
One important learning is to set up a correct experimentation process and document your process and learning. I use Airtable as it combines many tools like project management, data and also helps to build a proper database. So, it’s important to have a proper process and be able to see who’s doing what.
Secondly, it is to involve your other colleagues in the CRO process. Help them to achieve success with experimentation.
As regards failure, I’d say there aren’t any failures but only opportunities to learn. For example, I set up an A/B testing which resulted in a bug error and the developer team reached out to me asking how that could happen. As long as you learn, there are no failures.
I never start with the website itself. I want to fully understand the goals and business. If you are not aligned with the major goals of the company, then your efforts may not benefit the company.
So, I always start with understanding the business and what drives them. On the website I start with data. I run the first test as soon as possible for which I do a thorough research with various teams. I try to get a few links to get the organization enthusiastic.
The biggest is the product or the service itself. You can have the best website without any friction. But if it doesn't motivate the buyer, then it will not sell. It can be misaligned with the motivation or the problems faced by the customer. You need to look from the point of view of the visitor and develop a great product or service. However, you do need a great website to make the sale.
When I started with conversion optimization, there wasn’t much clarity or a well defined process to set up personalization. People lacked tools and resources. Few years ago, there started a lot of hype about personalization as if it was a goal in itself. However, it’s one of the means to an end. Those who have the right tools and resources are getting it right. Many companies are experimenting with it and I see it getting bigger and bigger.
Start with research about users, what are the problems they face, their obstacles, biggest data points of drop off. Start with thorough research, also scientific research and based on those insights, brainstorm test ideas. I generally work with various teams like designers, data researchers, psychologists. They all contribute to getting to those test ideas. With every experiment you can learn and come up with a follow up test.
At Online Dialogue, we have psychologists on our team and we are kind of famous for that. They help in analyzing and understanding and validating certain points- for example a user may say something and do another thing. We have seen that whenever we experiment with the help of psychologists, the win rate really increases. That combination is very strong.
I have an elaborate blog on this topic. You can read it here . Many people use theories and frameworks like Iceberg and PIE. The advantages with these are that they simplify the process a lot, however, the downside is that these are completely subjective and they are too much focused on ease. One-third of them is ease, that kind of kills the effectiveness of the experiment.
Some people use fact based models. It comes from data analysis, user motivation— these are a lot better but not evidence based. However, we like to use an evidence based model that helps us prioritize and conduct experiments in a goal oriented way.
If the motivation is higher, it gets higher prioritization and if it’s lesser, it gets lower prioritization. If you use Airtable, you always get the right data based on historical fact.
I mainly use the tools that my clients use. I use Airtable for documentation for all the processes, it sets up automation and saves time to conduct more experimentation. I use A/B testing, data analysis and personalization tools. I love the ones that democratize. It’s easy for colleagues to access data and use them.
I recommend my own courses on Udemy here and CXL blogs. Feel free to follow me on LinkedIn. Talk to fellow optimizers and learn from each other.
Be curious, have a growth mindset and never lose an opportunity to learn.