TL;DR: title

Problem statement:

My team is part of a larger consortium of semi-cooperative partners. I'd like to fix some communication gaps -- specifically, I'd like to teach everyone how to give each other a clue about *what* needs to be addressed when folks say they need a feature or there's a bug. So far, the list of bug reports is blessedly short. The real problem is the feature requests. I get statements like: "this output is hard to read" -- that is, an ask for help that I can't act on because there's not even the first clue about what a solution or better might look like. Having worked with this particular crew for a while, I refuse to allow a new feature that does not have clear exit criteria because those tasks become trojan horses for doing a lot of work for other teams that really they should have done themselves. My team doesn't have the domain expertise, so those tasks become a miserable exercise in mind reading and perpetually moving goal posts where any feedback is in the form of cryptic riddles from a capricious sphinx.

I'm thoroughly convinced that these peer organizations have real leadership problems, and as much as I'd love to go tackle those (I'm pretty certain they're the root causes), that's a no go. The only viable courses of action I see are to act "bottoms up" through shared culture at the IC level. The first obvious candidate for improving things there is addressing communication, because the bug reports / feature requests are only a symptom -- it's pretty clear that there are wildly diverging understandings of what needs to be done. I've had modest successes introducing these kinds of changes (I've managed to coach my team to reduce MR time to first feedback, time to resolution, etc. from weeks to 1-2 days). We made those changes by agreeing there was a problem, asking what was the absolute minimum that needed to be addressed, and then adding checklists for the minimums to the review. Have folks applied similar strategies to this kind of problem and, if so, any advice?

I found a few projects that had written guides for feature requests (I quite liked the RStudio docs on their GitHub). They all seem pretty reasonable and sensible, tho it's not obvious to me if they're "winning" / successful. Do folks have any experience with these? What was useful? Was there friction? What caused the friction?

​

Thanks!