Key Takeaways
- Run each step from the sidebar and clean the live sheet with no formulas.
- Standardise text first so fuzzy matching groups variants that exact lookups miss.
- Write results beside the source data, so the original values stay in place for review.
- The free tier runs once a day on up to 50 rows, and a pass raises the cap per run.
This method runs entirely in Google Sheets, so there is nothing to download. Open Extensions > Flookup Data Wrangler > Standard Data Cleaning, work through the steps below in order and keep the original values beside every result.
Why formulas fail on messy sheets
Built in dedupe and lookup formulas assume tidy values. Real sheets mix Acme Corp with Acme Corporation, pad cells with stray spaces and vary phone formats across rows, so exact comparisons miss pairs that a reader sees as obvious. Each helper column then adds upkeep, and one edited formula can silently change the result for the whole tab. The technique guides in remove duplicates and cleaning tips cover the formula path well. This post covers the other path: select the range, run each step from the sidebar and keep the reviewed result beside the original values. The same threshold advice carries over from fuzzy matching in Google Sheets.
Run Standard Data Cleaning from the sidebar
Open the sheet, select the cell where the result should start and open Extensions > Flookup Data Wrangler > Standard Data Cleaning. Pick a function, check its parameters and click Run. The result is written to the active range from the active cell, so you choose where the output lands instead of writing a formula that references it. The full list of functions, with parameters and examples for each, is in Standard Data Cleaning.
Nothing leaves the tab. The run reads the range you selected and writes the result from the cell you selected, so the sheet itself is the only place data is held. A run that takes too long can be cancelled, and the rows processed up to that point are still written.
Standardise, match and dedupe
Standardise text before matching. Suffixes, punctuation and case otherwise split groups that belong together, and the same threshold advice from fuzzy matching in Google Sheets applies here: start at 0.85 for names, raise it toward 0.90 when false positives appear and lower it toward 0.80 when expected pairs are missed. Match across one column or a whole table with exact, normalised, phonetic and fuzzy measures, then dedupe to one master row per entity and keep a note beside each merged row as the audit trail.
Run Find duplicates before the threshold is fixed.
It reports the groups and changes nothing, so the size of the problem is known before anything commits. The output is a review list: one row per duplicate group, keeping the first occurrence of each, written beside your data rather than into it. Then run Remove Duplicates once the groups look right, so one master row per entity survives and a note beside each merged row keeps the audit trail.
Place the results where you want them
Every function writes from the active cell, so the output position is your choice rather than a fixed rule. Select an empty range beside the source for a review copy, or a range whose values you intend to replace.
The sidebar warns before an active range will be overwritten, because writing starts at the active cell and continues across the range you selected.
The source range itself is never modified by a run. Keep the original column beside the result, compare the two and only then accept the cleaned values. That is the audit trail: the original and the result sit in the same tab, and a cancelled or partial run still writes the rows it finished.
Plans and row limits
The free tier runs once a day on up to 50 rows, which is enough to try every function on a representative sample. An Essential Pass runs up to 1,000 rows per run and an Extended Pass up to 10,000, with a 20,000 cell ceiling on each function. Profile Column stops at 5,000 rows and Find Duplicates at 2,000.
Run a smaller range first, review the output and only then commit the full tab. The limits and the reasoning behind them are listed in
Menu function limits.
When a sheet is too large or runs every day
Standard Data Cleaning is the interactive path: one range, one run, results you can see immediately. Two other cases sit outside it.
A job that repeats belongs in
Data Cleaning Automation, which runs the same cleaning on a schedule from the sidebar with pause, resume, run now, edit and clear controls. A dataset too large for one interactive run belongs there as well, because the scheduler works through the range in chunks instead of stopping at the per-run ceiling.
Frequently Asked Questions
Do I need formulas to clean the sheet?
No. The sidebar runs each step for you. Choose a range, pick a function such as Standardize Data or Remove Duplicates and the result is written to the cell you selected.
Where do cleaned values go?
To the active range, starting from the active cell. Your source data is untouched, so select an empty range beside it, or a range you intend to replace.
How large a range can I clean?
The free tier runs once a day on up to 50 rows. An Essential Pass runs up to 1,000 rows per run and an Extended Pass up to 10,000, with a 20,000 cell ceiling on each function.
Can I stop a run that is taking too long?
Yes. Cancel stops the job and writes the rows processed so far, so the partial result is kept rather than lost.