Must: The system must meet these requirements or is
considered a failure
Should: The system should meet these requirements, but if it doesn’t we can do it later
Could: The system could meet these requirements. No one will object, unless there are must and should requirements that are unmet
Won’t: The system won’t do this. It will make the system worse and/or any time spent on these things is completely wasted. Don’t do them.
Seems pretty straightforward. The differences are clear. Do them in that order. You don’t need any more information so get to work,
Not so fast. There are at least a couple of problems here. First, those are just labels. Labels on buckets of similarly important things. There’s no sequencing provided inside a bucket. What happens if there are more items in the must bucket than there are teams to work on them? Even if there’s enough time to serialize them, you don’t know which one should be done first. So it’s really categorization.
If there’s only one team, and the requirements are all completely orthogonal, sequencing doesn’t matter. Of course, in all the time I’ve been doing this I’ve never worked on a project like that. And I don’t know anyone who has. It’s probably happened somewhere, but it’s rare enough to not worry about right now. Which means sequencing is important.
Second, while those are words, not numbers, there’s really no difference between Must and Priority 1 (or 0, or -1). It’s just the group with the highest importance. And they both suffer from the same kind of inflation. Every group/team/stakeholder thinks their problem/requirement is the most important. Or if not critical overall, critical to them, so they label it must. Because we all know that the shoulds almost never happen and they coulds are there for amusement only.
Which is not to say that categorization is unimportant. It’s not. It’s critically important. But it’s not enough. You have to go beyond the categorization and really prioritize. You need an ordered list of what’s the most important, balancing urgency and short and long term gain. You need to keep that list current. And most importantly, you need to follow it. Even (especially?) when a single stakeholder starts arguing loudly for their favorite thing.