Ten years ago, in every software box there was a manual – even in the boxes of Nintendo games. Software developers used them to explain installation, hardware requirements, operation, troubleshooting and power user tips. I remember writing and rewriting them for my first company.

I haven’t seen one of these in years despite the fact that I try 10 new applications every day. These manuals simply don’t exist anymore.

Hardware requirements are moot. Most current phones and PCs run almost every app. Installation has changed to mean creating an account, achieved with one click if you use Facebook Connect, or an OpenID provider like Google.

And we learn new applications by (i) having defined conventions for where things are on a page; (ii) building simpler and more intuitive UIs with progressive feature discovery so users learn the app as they use it; (iii) fragmenting applications so that each is very good at just one thing.

As a result, to succeed new applications have to demonstrate they can accomplish one goal very quickly and easily. Only in this way will a user invest the time to learn the application.

After having grown a user base, a developer begins to add complexity to the application. In this way, as the application grows in-depth, the user base grows in understanding in parallel.

All these trends result in initial launches of applications which many call features, but are much better described as point solutions to solve user problems. By reducing the initial product to its minimal viable state, startups can quickly build a strong user base and eventually grow the first product into something much broader.

At the outset of launching an application, they keep the service simple and the value proposition crystal clear. If you build something useful, your conversion rates will be much better.