There are definite bad ways. I've watched a few companies go down the shitter using known bad ways.

I've watched a few companies go down the shitter using known good ways; and seen companies with mind-blowingly horrific software succeed based on marketing and contracts. There are a lot of anecdotes, but I have yet to see any data suggesting a correlation.

It also depends on your definition of successful.

To be successful does software need to (check off what you think would matter):

* Make money directly

* Make money in-directly

* Be used by at least 1 user (not including dev + dev friends and family)

* Be alive and working in 6 months

* Be alive and working in 2 years

* Be alive and working in 10 years

* Simply ship so dev gets a positive yearly review

* Be bug free

* Only crash N% of the time (who cares just re-run it and don't do the bad thing)

* Quality / features don't matter as long as marketing is talking about it and people play with it so it looks successful.

* Simply exist, but be subsidized by some one or something so devs keep hacking on it even if no one uses it.

* Still be useful when there are zero devs working on it.