About Excel Fuzzy Lookup
Match company names, product titles or people even when they are spelled slightly differently. Set a minimum similarity and get a score and the matched value for every row so you can review the results.
How to use Excel Fuzzy Lookup
- Upload the list to match and the reference list.
- Pick the columns and set the minimum similarity (80% is a good start).
- Run the lookup, review match scores, and download.
Limitations
- Everything runs in your browser’s memory, so the practical file size depends on your device. As a rough guide: up to ~50 MB or ~500,000 rows works well on a modern laptop; phones and tablets handle much less.
- Very large files can make the tab slow or crash. If that happens, split the file first or use a desktop computer.
- Outputs keep values and formulas’ results but not formatting, charts, images, pivot tables or macros.
- Fuzzy lookups compare every unmatched value against the lookup table, so very large tables (tens of thousands of rows on both sides) can take a while.
Frequently asked questions
How is similarity measured?
With edit (Levenshtein) distance: 100% means identical, 80% means about one character in five differs.
Exact matches are found first?
Yes, fuzzy matching only runs for values without an exact match.
It is slow on big files — why?
Each unmatched value is compared with every lookup value. Clean obvious differences first or split the file.
Disclaimer: Excel Tools by gumber.ai is a hobby project and is completely free to use. It is provided “as is”, without warranty of any kind. Always double-check results before relying on them, and keep a copy of your original files.
Related tools
Compare Excel Files
Spot changed cells, added rows and removed rows between two versions of a sheet.
Remove Duplicates
Find and remove duplicate rows based on all columns or just the ones you pick.
Excel VLOOKUP
Pull matching values from a second table into your sheet — no formulas, no #REF errors.
XLOOKUP / INDEX-MATCH
Look up any column in any direction and return one or many result columns.