A few years ago, I had a voice into the prioritization of work on an e-commerce system. Of course, we got more requests for functionality from internal teams than we could satisfy.

Initially, we tried to prioritize via a somewhat organic, conversational process…

Unsurprisingly, that just meant the most politically powerful person’s ideas always prevailed. Ok, so what?

It was impossible to offer explanations better than “X and/or Y said so.”

How could we better help folks not “in the room where it happened” understand the “why”?

I did a lot of reading and came across the concepts of “Cost of Delay” and “Weighted Shortest Job First” (h/t @jezhumble) in my reading, which seemed useful models for unpacking our beliefs about the relative “business value” of various efforts.

This exercise was incredibly illuminating and, with some, incredibly unpopular.

It turned out, a large fraction of “requirements”, even those with powerful political weight, did not pencil out in even the most optimistic cases.

We literally let people just make up the upside numbers, with the proviso that they should be prepared to defend them if asked.

Naively, I expected this information would be viewed as I viewed it—a welcome tool to help us make more sound business decisions, which was the language these same people kept saying was the important aspect of their perspective (vs engineers).

Here’s what I learned:

Structured models and frameworks do help you make better business decisions.

But that improvement necessarily undermines the political power of folks who have acquired their influence via some path other than sound decision making.

Those folks will fight tooth and nail to undermine any kind of scrutiny or rigor that threatens their ability to continue to make arbitrary “gut feel” decisions.

They will often claim that their decisions are data-driven and/or beyond the capacity to understand of those outside their discipline.

Every single time since, when I have finally seen behind the curtain, there is, in fact, no data and no rigor. Just pithy aphorisms and “it depends” hand-waving.

No model is perfect. Some things are hard to quantify. But if your “business” decision making doesn’t have some kind of structure beyond “a few people got in a room and agreed” which can explain how you chose between various options, you’re making a lot of bad decisions.