For four decades, Excel enforced a rigid contract: one cell, one value. Every workaround — comma-separated strings, helper columns, TEXTJOIN hacks — was a user rebellion against that constraint. Microsoft is now capitulating with lists, arrays in cells, and nested arrays, rolling out to Windows and Mac beta channels. Lists let you store multiple discrete values in a single cell while keeping them individually addressable for filtering and calculation. You create one with Ctrl+J or Insert > List, separating items by commas or semicolons. Referencing a list cell spills its contents into adjacent cells, bridging the old single-value world with the new multi-value one. Filtering now works on individual items within a cell, not just the full text blob. Arrays in cells go further. Any spilling formula can now be wrapped in braces to collapse its output into a single cell. This is a fundamental change to Excel's calculation model: dynamic arrays, introduced a few years ago, forced results to spill across the grid. Now you choose. Nested arrays — arrays containing arrays — eliminate the old #CALC! errors and truncated results that plagued complex formulas like TEXTSPLIT across rows. Four new functions ship alongside: FLATTEN removes nesting layers from arrays, while HAS, HASANY, and HASALL provide membership testing. These are utility functions, not headline features, but they close the gap between "arrays exist in cells" and "you can actually do useful things with them." The run-tracker example Microsoft demos — split times per kilometer stored as a per-row array with inline stats — shows the target use case: tabular data where row length varies. The compatibility tax is real. Nested array calculations require a new Compatibility Version 3 setting per workbook, and switching may change existing formula results. The known limitations list is long: conditional formatting, data validation, charts, PivotTables, and Power Query all fail to inspect or use array contents properly. This is not a finished feature — it is scaffolding. Microsoft is making a generative bet here. The constraint being lifted was genuinely limiting; anyone who has built a serious spreadsheet knows the pain. But the execution risk is high. Excel's ecosystem — add-ins, corporate templates, regulatory spreadsheets, Power Query pipelines — depends on the one-cell-one-value assumption. Breaking that contract in ways that are invisible (Compatibility Version 3 silently changing formula results) creates fragility that won't surface until production workbooks break. The 20-year question is whether this becomes Excel's version of structured data types — a genuine paradigm shift that makes spreadsheets more capable — or whether the compatibility overhead and integration gaps keep it a power-user curiosity. Microsoft controls the platform, so the bet is theirs to win or lose. But shipping with six major subsystems unable to read the new data type suggests the internal coordination cost is already high.
