1. Instead a BDD format, it uses markdown specification-style language. This fits well for end-to-end user-journeys, instead of a strict Given-When-Then which is more suited for very specific test-case style of testing.
2. If you have data-heavy testing, Gauge helps. It allows you to manage your data in files, apart from tables, and gives an out-of-the-box way to read from these files, thus making data management much simpler and neater.
3. Gauge is built with an architecture such that you have a core library, and then there are plugins to it. Every language binding is also a plugin. So, all the core features are in the core lib. Thus, if you change tech stack, and move from say C# to Java, the features you had would mostly be there ready for you. Thus a smaller learning curve (unlike Cucumber which is re-written from scratch everytime you change tech stacks).
4. Gauge also comes with out-of-the-box parallel running mechanism. It just needs a simple command, specifying the number of threads you want to run them with, and that's it. No additional plugins required.
Hope that helps.
