So if I’m reading this right, what you’re saying is that the content bloat exists so that existing engagement metrics look better? That tracks. The idea that they’re grasping entirely wrong facts instead of “just” misunderstanding the dimensions’ causal relations, didn’t occur to me, but given that it’s a simpler and still just as plausible explanation, Occam’s Razor suggests it’s more likely. In fact, I’d bet it’s (more) correct.

In my naive position as BI developer, I tend to insist my “customers”/consumers focus on clarifying the questions they want to answer, so that I can find the appropriate measures to calculate, then work on narrowing down the disconnect between what data I have and what they want to know.

Of course, the situation “my consumers have latched on to this specific measure, how do I explain that it doesn’t actually answer the question they’re asking?” and the related problem of “how do I explain that the question they’re asking isn’t quite as simple?” occur often enough, but between the faith in my abilities and understanding of our data (on which I’ve become the de facto authority) and my own practice at picking examples that show such disconnects, it’s usually little more than a routine exercise.

But most of my consumers (or at least the contacts for my models and reports) are operative or lower management that have either been relying on my service for years now or come to me after word-of-mouth from other people that have, so the bullheadedness of C-level executives just grabbing a single, intuitive measure and holding on to that - actual semantics be damned - is far from my daily business.

Yeah, that one is on my To-Read-List still. I’ll move it up.