The Problem with Personalization in Ecommerce: Correlation vs Causation

– Ecommerce industry lacks understanding of correlation and causation in relation to personalization and conversion rates.
– Quizzes and subscription status may indicate higher intent to purchase, but they do not directly cause conversions.
– Personalization and specific marketing channels do not significantly drive conversion rates; understanding the customer journey and improving all aspects of it is more effective.Focus on understanding the customer journey and improving all aspects of it to determine the data combinations that lead to the highest conversion rates.E-commerce as an industry has a big problem of not understanding correlation and causation. This came up in a chat today again around personalization. Quiz companies keep hyping that they have high conversion rates and increased revenue due to personalization. Quizzes aren’t causing the conversion. At best, the quiz serves as an indicator that someone is more active in the buying process and already has more intent than a typical customer. The correlation is that the person is further down the customer journey than someone who is just browsing. Meaning that you can draw a correlation, but not causation.

In fact, we see this within our own data. Subscribed customers have a conversion rate for us of about 30%. They didn’t convert because they were subscribed, but there is a correlation to a higher conversion rate due to the higher intent.

Whenever I take a quiz, I go back and change answers to see if there are different results as I’m often between a few things. This is actually where most people fit in. They are still trying to figure things out. I’m actively looking to find out what might be a good fit for me, so because of this, I’m more likely to convert. This actually creates some pretty dirty data 😉 It’s not the personalization that’s making it happen, it’s the place where I am in the customer journey that makes me more likely to convert. BUT THAT DOESN’T FIT INTO A NARRATIVE OF SOMEONE SELLING THE SOFTWARE WHICH IS BEING MEASURED ON ROI TO JUSTIFY THE COST.

The same goes for email being responsible for revenue. It’s not in any way other than a reminder for the right time. There was already intent to purchase by virtue of signing up. The same goes for SMS with that 30-day attribution window. The intent is already there to purchase at sign up. Correlation, not causation again. If these things were really driving conversion rates like people claim, then the overall conversion rates on the website would go up. In fact, you can send out pretty much anything and you’ll get sales. You can try it, just have a CTA for them to get back to the website. People will purchase if the timing is right.

So maybe, just maybe, we can take a breath and start actually looking at the practical logical reasons for increased conversion. Hint: it’s not software or a channel. It’s understanding the customer journey. I’m a guy that helps businesses collect and leverage zero party data telling you that in most cases a lot of personalization is a waste of time.It’s instead a macro approach to driving improvement.Improve all aspects of the customer journey to understand what data combinations lead to the most conversion and talk to that audience. Build a customer journey for that audience.#ecommerce #customerjourney #marketing

Leave a Comment

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

Scroll to Top