The user's question is sensible and direct, and it points to a real gap in how traditional spreadsheets handle data at scale. They want a file-based data source that refreshes all pivot tables in one action, supports calculated fields for weighted averages, and performs well on hundreds of columns and hundreds of thousands of rows. That's a reasonable set of requirements, and the fact that it feels like a wish list tells you something about the current state of the tools.
Here is the core problem: most spreadsheet applications treat pivot tables as static snapshots tied to a fixed range or a named table. When the source file changes, new rows, new columns, recalculated fields, you are often left manually updating each pivot table's source range. The Refresh All button is supposed to solve this, but in practice it only works if the data structure hasn't shifted. If your source file adds a column or changes a formula, Refresh All becomes a partial solution at best. The user's demand for a single refresh that updates every pivot from a changed source is not a nice-to-have; it is the baseline for any data workflow that evolves over time.
Calculated fields add another layer of friction. Weighted averages are common in finance, operations, and analytics, yet most spreadsheet engines force you to compute them outside the pivot table or rely on helper columns that break when the source updates. A data source that natively supports calculated fields is not exotic, it is what anyone building a real reporting system should expect. The fact that users have to ask for it as a special feature shows how far legacy tools have drifted from actual user needs.
Performance is the third rail. Hundreds of columns and hundreds of thousands of rows is not an extreme dataset by modern standards, yet it chokes many spreadsheet engines. The user is not asking for a database; they are asking for a file-based tool that does not fall over at moderate scale. That is a reasonable bar, and it is one that AI-native spreadsheet technology clears easily by processing data in memory with columnar efficiency rather than recalculating cell-by-cell.
Our take is this: the user's criteria are not ambitious. They are the minimum viable requirements for anyone who treats spreadsheets as a serious data tool. If your current solution makes you hesitate before adding a column or recalculating a field, the problem is not your data, it is the tool. Explore options that treat files as live, refreshable sources with built-in calculation support. That is where the practical transformation begins.