This is a story about expertise hitting a wall, not a story about a spreadsheet failing. The person who wrote this knows their data, their workflow, and their boss's expectations inside out. They built a solution that worked, manually, carefully, month by month. The problem isn't their logic. It's the tool they're forcing that logic into.
When a single formula change triggers a twenty-minute calculation thread, the spreadsheet has stopped being a tool and become a bottleneck. That lag isn't a sign of complexity done right; it's a sign that the spreadsheet is being asked to do something it was never designed for. Excel is a powerful canvas, but it was built for static analysis, not live operational data flowing in from multiple forms. What this person needs is a system that separates the data from the calculations, so the computer isn't recalculating every cell every time one entry changes.
The practical path forward is to stop treating the Excel workbook as both the database and the report. Keep the form data in its own clean table, no formulas, no formatting, just rows of raw inputs. Then build a separate reporting layer that pulls from that table, calculates the averages and percentages, and updates on demand or on a schedule. That separation is what modern data tools do naturally. It's why an AI-native spreadsheet can handle thousands of rows without choking: because it doesn't recalculate everything every time. It only recalculates what changed.
The boss wants this to work indefinitely. That's a fair ask. But the answer isn't to optimize a workbook that was already stretched past its limit. The answer is to move the heavy lifting to a tool that respects the difference between storing data and analyzing it. This person has the expertise. They just need the right architecture to let that expertise run free.