A Reddit user shares a screenshot of two pivot tables and asks for help calculating a true average instead of the average of averages. This is not a minor formatting issue. It is the exact moment when a tool stops serving the user and the user starts serving the tool.
The problem is straightforward. The pivot table automatically computes an average per group, then averages those group averages. That gives a misleading result unless every group has the same number of data points. The user knows this. They want the overall average, the one that accounts for every row. They cannot get it from the built-in pivot table summary without manual workarounds, adding a calculated field, writing a formula outside the table, or accepting the distortion. This is not a skill gap. It is a design limitation that forces users to become experts in the tool's quirks just to get an accurate number.
Think about what that means in practice. Someone who has already invested time learning pivot tables, who has a clear analytical question, hits a wall not because the question is hard but because the tool's default behavior assumes a specific data structure. The user is not asking for a complex statistical transformation. They are asking for a correct arithmetic mean. The tool should handle that out of the box. When it does not, the user spends time searching forums and posting screenshots, all to compensate for a feature that was designed for convenience at the expense of accuracy.
This is the kind of friction that traditional spreadsheets have normalized. Users accept that they must learn workarounds, that the tool will occasionally produce wrong answers if they are not careful, and that the burden of correctness falls on them. An AI-native spreadsheet should reverse that. It should understand what the user intends, a simple summary of the real average, and produce it without requiring the user to outthink the software. The tool should adapt to the user's data, not the other way around.
The lesson here is not that pivot tables are useless. They are powerful for many tasks. But the automatic average feature reveals a deeper problem: too many spreadsheet tools prioritize speed of generation over correctness of result. A user should never have to ask how to get a true average. They should only have to ask what they want to learn from their data.