The user's question reveals a common frustration: they want to preview which entries a fuzzy match *would* fix, without committing to the match itself. Their proposed workaround, running the fuzzy match in Power Query, then checking if the result matches the original, is functional but inelegant, especially when juggling eight separate tables. We think there's a cleaner path, and it starts by rethinking the problem.
The core issue isn't the fuzzy match tool; it's the need to surface *near-matches* across datasets without triggering the full transformation. A better approach is to pre-process each table by normalizing the entries into a consistent format before any comparison. For stock tickers like "META" versus "META US," that might mean stripping suffixes, removing spaces, or converting to uppercase. Once each table has a "cleaned" column, you can use a simple exact match on that column to find entries that *would* have matched under a fuzzy rule. This lets you flag candidates without running the fuzzy algorithm at all. The manual review you mentioned still happens, but now you're reviewing a much smaller, targeted list.
For the eight-table challenge, don't compare every table against every other table, that creates combinatorial complexity. Instead, designate one table as your master reference. Clean it, then run a lookup against each of the other seven using your normalized column. Any entry in a secondary table that doesn't exactly match the master's cleaned version becomes a candidate. This gives you a single, organized list of potential mismatches across all sources. It's not a magic bullet, but it's simpler than stacking fuzzy matches and checking results.
The practical takeaway: you don't need to run the fuzzy match to know where it would help. Build a lightweight normalization step first, then use exact matching to isolate the problem entries. That approach scales across many tables, keeps manual review manageable, and avoids the computational overhead of fuzzy logic until you're ready to apply it. The goal isn't to eliminate human judgment, it's to point that judgment at the right rows.