If you've ever spent two hours wrestling with a government website to get a simple table into Excel, you know the specific frustration that user described. Their post is a perfect snapshot of a problem that has nothing to do with inexperience and everything to do with how traditional spreadsheets handle the modern web. The FDA's drug approval data is public, valuable, and deliberately structured for human reading, but copy-paste collapses it into a single cell, and Power Query chokes on the page's layout. The user blamed themselves. They shouldn't have.
This is the moment where the gap between "data exists" and "data is usable" becomes painfully visible. Legacy spreadsheet tools were designed for a world where data came in clean rows and columns, not nested HTML tables buried inside government portals. The user tried the standard workarounds: copy-paste, then Get Data, then asking for help. Each failure reinforces the idea that the problem is theirs to solve. But the real issue is that the tool they're using treats every import as a manual extraction project. It doesn't see a table on a webpage as a living data source, it sees a formatting puzzle.
What this means practically is that researchers, analysts, and anyone who needs public data loses time they don't have. A two-hour struggle to import one table is two hours of friction that could have been spent analyzing trends in drug approvals, checking for patterns, or moving the project forward. Worse, that friction discourages people from working with public data at all. When the barrier to entry is a multi-step, error-prone process, the data stays on the website, not in the spreadsheet where it can drive decisions.
The solution isn't for users to learn more Excel tricks. It's for the spreadsheet itself to understand what it's looking at. An AI-native spreadsheet can recognize the structure of a table on a page, extract it cleanly, and present it as a usable dataset, no copy-paste gymnastics, no Power Query configuration. The FDA's drug approval table should be three clicks away, not a two-hour ordeal. That's not a luxury feature; it's a baseline expectation for a tool that claims to handle modern data work. The user who posted that question deserves a tool that meets them where they are, not one that makes them doubt their own competence.