Copy-paste formulas to dedupe a lead list in Google Sheets
Paste a formula into row 2, review the matches, then drag down.
Normalize a name
A: raw company name
=GPT("Return a normalized form of this company name: lowercase, no punctuation, drop legal suffixes (inc, llc, ltd). Name: " & A2)
Same entity?
A: name 1 · B: name 2
=GPT("Are these the same company? Answer yes, no, or unsure, then a confidence 0-100 and a short reason. A: " & A2 & " | B: " & B2)
Pick the cleaner record
A: record 1 · B: record 2
=GPT("Which record is more complete and correctly formatted? Return 1 or 2 and a short reason. 1: " & A2 & " | 2: " & B2)
Short answer
Deduping a lead list in Sheets means finding records that are the same entity despite different spellings, suffixes, or formats, then keeping one clean version. GPT for Sheets normalizes names and judges fuzzy matches that exact-match formulas miss.
Fastest path: Install GPT for Sheets → add a normalized-name column → use the match formula on candidate pairs → review → resolve.
This page is for RevOps and marketing ops cleaning imported or merged lists. Use AI matching as a reviewer that proposes duplicates; confirm before deleting, since automated dedupe can merge distinct records.
Workflow
A practical sheet for this workflow usually has these columns:
| Column | What to put there | Why it matters |
|---|---|---|
| A | Raw company/contact | Row anchor |
| B | Normalized name | Strips suffixes/case for comparison |
| C | Domain | Strong secondary match key |
| D | Duplicate-of pointer | Links a row to its primary |
| E | Match confidence | How sure the match is |
| F | Keep / merge flag | Human decision column |
Normalize before you match
Exact matching fails on ‘Acme, Inc.’ vs ‘Acme Incorporated’. Use a normalize formula to lowercase, strip punctuation and legal suffixes, and standardize spacing so obvious duplicates line up for sorting and review.
Judge fuzzy pairs, then resolve
For pairs that still differ, ask the model whether they are the same entity and how confident it is. Sort by confidence, review the borderline cases, and only merge after a human confirms in the keep/merge column.
Use cases
- Import cleanup: catch dupes after a CSV merge.
- CRM hygiene: reduce duplicate accounts.
- List merges: reconcile two sources.
- Routing accuracy: avoid double-touching a company.
- QA: surface borderline matches for review.
Best for / not best for
Best for: RevOps and marketing ops who clean lists in Google Sheets and want help catching fuzzy duplicates that exact-match formulas miss.
Not best for: fully automated merging without review, or compliance-critical record-linkage requiring deterministic, audited matching; confirm matches before deleting data.
The strongest use case is normalizing names to catch obvious duplicates, then using AI to judge only the ambiguous pairs, with a human approving every merge.
Internal links and next workflows
- GPT for Sheets product page
- GPT for Sheets pricing
- AI lead enrichment guide
- Categorize leads by industry
- Apollo export enrichment
- B2B sales account research
Safety, compliance, and data quality
Automated dedupe can merge distinct records, so treat AI matches as proposals. Keep a human keep/merge decision, back up data before bulk changes, and use deterministic rules for compliance-critical record linkage.
Frequently Asked Questions
Will it delete duplicates for me?
No. The formulas propose matches and a confidence score; you keep a keep/merge column and confirm before removing anything, so you stay in control of the data.
Why normalize before using AI?
Normalization catches the easy, high-volume duplicates cheaply and deterministically, leaving only the genuinely ambiguous pairs for the AI to judge.
Can it match on domain too?
Yes. Keep a domain column as a strong secondary key; matching on normalized name plus domain is more reliable than name alone.
Is the matching guaranteed correct?
No. It is a reviewer that proposes matches with confidence; for compliance-critical linkage use deterministic rules and human confirmation.
Start cleaning your lead list in Google Sheets
Add a normalized-name column, install GPT for Sheets, and let AI surface the fuzzy duplicates while you approve every merge.