Problems with Micro-frontends

For the past six months, our team has been rewriting our application using a micro-frontend approach. Given how popular micro-frontends appear to be at the moment, I wanted to share our experiences and the lessons we learned.

Micro-frontends can be a fantastic tool, but they are not a silver bullet. Depending on the structure of your project, team and business, you may not be able to take advantage of the benefits. Worse, micro-frontends may undermine the architecture of your application and hinder your team’s ability to deliver. For our team, micro-frontends proved to be a particularly poor fit and an example of where the micro-frontend approach isn’t the best option.

The (potential) benefits of micro-frontends

This experience happened to be my second encounter with taking a micro-frontend approach. At a previous role, we had a team maintaining a shell Android application and four teams, each responsible for a separate webview within the shell application. Each team addressed an entirely separate business concern, and the autonomy micro-frontends provided us ensured that each team could own the entirety of their stack. Teams were able to develop and release independently and didn’t have to worry about affecting each other unless they touched the shared library or shared workflows. Maintaining the shell and the…