Well written! I agree with some but not all of this.

I think the MVP has the minimum amount of features that are important to test hypotheses and learn from. There may be other features that people expect in an app but aren’t truly important to test.

For instance, you mention search functionality in a new social networking app. While this is an expected feature, I don’t think it’s really important to have this in an MVP that’s being tested with early adopters. A search bar is not going to differentiate you from Facebook, Twitter, etc., and it may take days to weeks to implement search functionality.

Rather, the focus should be placed on the features that are going to differentiate your social network from others. My guess is that Snapchat did not have a search bar in their MVP, and they focused more on the video capture aspect of their app.

For the B2B SaaS MVP that I’m building, I’m actually going to leave out certain user management and administration features. These are certainly features that are eventually necessary. But everyone knows what these flows should look like, so they won’t be a differentiator and I won’t learn anything from including them. And they’d likely take a few weeks to build. So I’d rather spend time managing those features manually in order to get the MVP out earlier so I can learn faster.

Just my $0.02!