Certainly, the upside is larger - cannot argue. The challenge, in my mind, with respect to software is the unforeseen, not-entirely-in-your-control challenge o

Certainly, the upside is larger - cannot argue. The challenge, in my mind, with respect to software is the unforeseen, not-entirely-in-your-control challenge of integration, bug-fixing, etc. I'd much rather have delivered 95% of the scoped work, tell the client essentially, "we can do X, but it could require some non-trivial expenditure on your part or, since we've done a great job setting priorities to date, maybe that feature is not mission critical." Time and materials gives both parties some agency. Of course, if your powers of predictions are infallible, then perhaps you're better off with fixed scope! (My powers of prediction are infallible too...until they aren't!)