I once worked on a project where the goal was to improve the reliability of MS Teams. The skip level manager was not happy when we said we cannot estimate what the revenue gain would be because it can't be measured.
This risks getting into the Fisherian/Bayesian stats wars again, but it would be possible to project a revenue gain, even in advance of making the change. You'd do this in terms of improved customer retention, lower cost of future customer acquisition, increased price a Teams license could sustain based on a rep for higher reliability, etc.
Obviously it's not a change that you'd be able to pin a specific human decision-maker (who was marginal on Teams) down as to this change being the difference between a sale or not, but nor is the change in revenue going to be a random number uniformly distributed in (-∞,∞).
If the change was revenue-neutral, the skip level would probably have been justified in seeing if the teams working that project could have found something to do customers actually care about instead.
How do you estimate the improved customer retention and improved future customer acquisition of something as nebulous as, "this does very little on its own but is a tiny part of making Microsoft software feel more robust in general"?
There's probably various methods, but if you don't already have quantitative estimates, you'll probably end up polling SMEs you'd trust for this on how much increase in retention and decrease in acquisition costs you'd expect for a notational more robust Microsoft. Average those out.
Even 5 people is enough to significantly narrow the uncertainty (at least, as compared to the prior hypothesis that the true change in odds is somewhere from -1 to 1).
And like a good spam filter, you don't need to be perfect, you just need to be able to turn your initiatives into a forecast of future changes if implemented, and then measure what actually does change for those initiatives you do implement. Even simple linear regression models can be useful in tracking results compared to forecasts.
Most people would probably expect that the improvement is close enough to zero to be nearly indistinguishable from it. And this is why most enterprises don't invest much in reliability work like this, because it doesn't actually seem to move the needle for customer purchasing.
Any manager who doesn't push back when given a perverse incentive should be fired. They are not leaders, they are followers... And not even very good followers.
Typically from C-levels who offer intuitive seeming, but internally inconsistent goals. For example, "We need to increase sales and decrease spending" might be two inconsistent goals if the only way to increase sales is to increase spending.
A bad manager will accept these inconsistent, and mutually exclusive, goals as if handed down on high by the gods and find counter-productive ways to save money or increase sales - even at the expense of actual profit. There is always a way, if you don't care about the actual outcome. These people will Goodhart the company into a terrible position, as long as it means they meet their arbitrary KPI's and get their full bonus.
A good manager might instead gather some data, then come back to the executive the next day with a data backed explanation showing why meeting both goals might be possible, but still wouldn't be advisable due to the negative externalities. A good executive will actually listen, because it's backed with real data they didn't have when they made the call.
Unfortunately if either the manager OR the executive are bad or thinking with their ego's... The entire thing falls apart. Which is why it's so very common.
