Any reason for not releasing an MVP with frequent follow-on releases instead of a big-bang release? I ask because I suffer from the problem of not releasing often enough. Saw matching symptoms in your post that made me wonder what your reasons were.
I'm rewriting an old, limited BASIC (!) application. So my "customer" knows quite well what he needs (he needs the same, but better). So requirements are rather easy to grasp. I also demonstrate progress every now and then so that he can get a good feeling of what I do. Now, the code has become good enough that I can give it to him to test it (and he'll behave as the usual customer : he will not test enough :-) ). Moreover, I have access to its current database => i've got a lot of data to validate the business rules side of the program.
Concerning the big bang now... The program is quite a "whole" so it's pretty hard to release a part of it. I basiacally made two parts : the classical ERP stuff and the time reporting (machines/humans). Also, working for a small, privately owned company, to replace an existing software is different than a brand new "super idea" web site. That is, the room for error is super small. And since it's a side project, I cannot provide on site support within 8 hours. Therefore, I think the big bang is the only option (well, for both parts of the program). As said, I'm working on a stable requirements basis => I'm confident I can manage the application development and code quality with standard testing...
In retrospect, I think the project is nice but a bit too big. 2 years, day in day out, is quite an endeavour. And although I like it, I feel it has to finish now...
In case you wonder : I didn't use an existing ERP solution because after some tests by my brother-in-law-customer, it appeared that they are either too expensive, too complicated/slow or they impose too much of an administrative burden (eg. ned to declare every single stocks before being able to actually write some production orders...) Hopefully, what we're doing will prove more efficient for our case (and, let's dream, there might be other people interested, I mean, rich people :-)))
On this concept, does anyone know if there's a model for early release of a mobile app, or does that pretty much have to be a big-bang release? My app will probably be freemium with many social options and tie-ins, so I'm not sure how to "release early" with something like that.
