Research and key learnings

I started my process by reviewing documents provided by my partners and chatting with other UX designers who worked closely with the Amazon personalization algorithm on device. I used the key learnings below to guide my designs and recommendations.

Video content is the highest use case at 28% for customers on Fire tablet devices.

Showing multiple apps at once can lead to higher engagement.

Previous feedback from customers indicates that recommendations can be inaccurate or unhelpful at times.

No access to 3rd party content/data for our customers.

For the proof of concept, I focused on representing 1st party video content from the IMDb TV app because of the content restrictions the team faced for 3rd party apps. Eventually, the product would support recommendations for all apps and games as well as books.

I had to account for the potential edge case of customers being offline when viewing the ads. So, I started by creating simple user flows and wireframes to better define that experience.

Product design

One of the challenges when creating a template-based product like this is ensuring high-quality design and reducing design fatigue. My concept generated a random background for each ad from a library of subtle patterns. I also knew that we could easily automate a gradient overlay based on primary colors in the featured image.

I worked through many different designs and iterations before landing on my final product design, which featured an interactive carousel of three video titles.

Company
Amazon