Say you took the last post to heart. You asked broadly, you collected ideas from everyone who touches the work, and now you have a list of thirty things people would hand to AI. Progress. Except the list has to become a plan, and a plan means choosing — and choosing is where it quietly goes wrong.
Watch what happens next in almost every company. The ideas get discussed in a meeting. Someone with a strong voice and a good relationship with the boss makes a confident case for their idea. Heads nod. It goes to the top of the list. A quieter person's idea — which happens to save four times the hours — gets a polite "let's park that" because they described it in one hesitant sentence and moved on. The backlog didn't get prioritized. It got campaigned.
The bias with a name
This isn't a character flaw in your team; it's one of the most robust findings in the study of group decisions. Analytics leaders have a nickname for it: the HiPPO effect — the Highest Paid Person's Opinion — the tendency for a group to defer to the most senior or most forceful person in the room rather than to the evidence. The term was popularized in the analytics world precisely because so many data-informed decisions get quietly overridden by whoever outranks the data.
Underneath the nickname is decades of research. Classic experiments on group discussion found that teams disproportionately spend their time on information everyone already shares, while the unique knowledge held by one person — often the most decision-relevant knowledge — never makes it into the conversation at all. Groups, left to talk freely, converge on the comfortable and the common. The four-times-better idea dies not because it lost an argument, but because it was never fully heard.
An unstructured meeting doesn't surface the best idea. It surfaces the best-sponsored idea — and those are almost never the same thing.
Why "it feels important" isn't a number
Here's the deeper issue. When ideas compete on advocacy instead of arithmetic, there's no shared way to compare them. "This would really help the sales team" and "this would save me two hours a week" and "this is strategic" are three different currencies, and a meeting has no exchange rate between them. So the tiebreaker defaults to confidence and volume — which is to say, to whoever is most comfortable taking up airtime.
Two things almost never get stated for each idea, and their absence is the whole problem:
- Time saved. Not "this is valuable" — how many hours a week, across how many people, does this actually give back? An idea that saves one person ten minutes and an idea that saves twelve people an hour look identical on a whiteboard until someone does the multiplication.
- Effort to build. Not "we could probably do this" — is this a prompt someone writes this afternoon, or a six-week integration with three dependencies? Ideas with invisible effort get championed as if they were free.
Without both numbers, every idea is just a story competing with other stories, and stories are won by storytellers. This is exactly why so many AI initiatives stall after the enthusiastic kickoff: the effort of picking well is real, and unstructured groups are demonstrably bad at it. Behavioral research on decision-making is blunt about how much noise and bias creeps into expert judgment when it isn't disciplined by structure — and a backlog debated in a room with no shared units is that noise, made into a roadmap.
The cost of getting this wrong
The popularity contest doesn't just pick suboptimal projects. It teaches your team a lesson: that being right doesn't matter as much as being loud. The quiet person with the four-times-better idea learns not to bother next time. You lose the very input you worked to collect — and worse, the people you'd most want to hear from are usually the ones least inclined to fight for airtime.
So the problem isn't that you don't have ideas. It's that you have no honest, shared way to rank them, and the default ranking mechanism — who argues best — is actively hostile to good decisions. In the next post we'll look at the framework most teams reach for to fix this, the impact-effort matrix, and why even that right instinct usually gets used in a way that quietly reintroduces every problem it was supposed to solve.