Ask HN: Non tech founders, how do you assess the quality of your software?

As a non-technical founder without enough resources and work to justify hiring your own software engineers, entrusting an agency or freelancers to build the software powering your startup; how do you assess whether you are getting your money worth?

Specifically:

1. How do you make sure that the software isn't a giant unmaintainable spaghetti mess that will cripple you later as you grow and want to add features?

2. How do you assess whether the software will be easy to extend and maintain in the future, not necessarily using the same people/agency who built it?

3. How do you tell if the stack choices the agency/freelancers made are the right ones for your application and the future?

4. Even though it works well and looks good in the demo, how do you tell if it will go the distance or constantly crash and be down?

5. How can you tell how expensive it will be to run? I have heard and read that AWS can get very expensive, very fast.

6. How important is all the above altogether? Is it acceptable that your first version is pretty much a throw-away and that you'll rebuild it all from the ground up once you have your own software team? What if you don't plan on getting a software team at all (not a tech company at the core) but still want dependable software?

Any experience you could share on the above (both good and bad)?

Thanks!