The AI tool gives you an answer. Knowing whether it's the right answer is a different skill
The Setup
The CPG data world changed faster than its people did. When COVID emptied the offices, companies kept hiring remotely for analyst roles that used to be trained in the hallway, over lunch, at the desk of a veteran who had been at it for fifteen years. Those veterans weren't there. So the new hires learned the platforms. But not the judgment.
The AI Issue
AI tools have made this worse, not better. When a junior analyst used to make an error, it looked like an error: a wrong formula, an off number. Now AI produces confident, well-formatted, plausible outputs that are contextually wrong in ways that require experience to catch. The output looks like the answer. It isn't always.
The Three Part Gap
Platform Knowledge
How to pull the data
Category Knowledge
What the data should say
Judgment
When to trust the data and when to push back
Most training stops at Platform Knowledge. Most AI tools accelerate Platform Knowledge. Almost nobody is teaching Judgment.
The Stakes
Bad outputs don't stay in the analyst's data sheets. They go into decks. They go to buyers. They inform promotion, distribution, and assortment decisions. One wrong read on a velocity metric can cost a brand a shelf placement. The math is unforgiving.