Optimize globally (end of the journey) or locally (next step)?

Gwendal

This is a question I’ve been asking myself for a while and we now have some results to share!

We’re looking at a common scenario: Visitors land a home page or a landing page with a Call to Action for registration. They can use the solution and then buy the product/service.

Should you optimize your page for Registrations (hoping that more registration will lead to more purchases) or optimize for the end goal: purchases? Does that make a difference?

We ran an intelligent experimentation with 2 variants on a home page messaging:

  • One that focuses on value and expertise
  • One that is upfront with the price

We are running Cauzal on the 2 variants with the optimization target being Purchase (starting 9/16), and also measuring Registrations. (Note: Cauzal is NOT A/B testing, it displays the content to visitors based on model built with machine learning, which ensures that all variants find their “sweet spot”, and no loss of conversions is expected)

Now, let’s look at the performance of both

Purchases - Performance
Leads - Performance

What’s interesting here?

  1. Both variants are going up, and that’s what any conversion optimization expert should expect of their experimentation solution.
  2. When the optimization for Purchase is activated on 9/16, we see a clear trend change for purchases, with an increase in the order of 15–20%. This means that changing the landing page DOES impact the purchase steps later
  3. Registrations are slightly increasing but not at the same rate: visitors are served the variant that will maximize the change of purchase, not only registration and we can clearly see that the impact is not the same (Sensitivity insights is a great way to understand that: check this article for more on this topic)