[omarhaneef]

I don't develop web apps for a living but even I would have thought the obvious solution is to first get v1.0 up in Rails/Django or whatever you know and then try new frameworks etc.

Actually, if you want it to be a business, the first step would be just use something else out there.

Buy > Customize > Build

It would have been interesting to see what his wife uses at the moment. I think he could have given her functioning software by just taking her excel sheets (for instance, if that is what she used) and just put them online with Google, and connected them with Scripts and an interface with Forms.

After that was working pretty well, maybe he could offer to set it up for some others.

Only after that would he selectively take pieces of the system and change it to Django. So first, have the forms in Django talk to their sheets. Then have their sheets dump into his reporting system etc.

[karmakaze]

The "Buy > Customize > Build" logic holds for supporting software. When it comes to the core of your business, Buy is usually not an option and also doesn't give you a competitive advantage. Also the effort of Customize can often be greater than the Build path in short order. I've experienced this with one startup that thought that it made sense to customize a Drupal CMS for a bespoke experience with hybrid human/ML curation.