Identify the fields.
Email, phone, names, domains, addresses, CRM IDs and dates are given different roles. You can correct the mapping before rescanning.
Free CRM data utility · Browser-only
ONE CUSTOMER.
NOT THREE RECORDS.
Find duplicate contacts or companies hiding behind spelling changes, missing fields and messy exports. Review the evidence, choose the master record, resolve conflicts and leave with a clean merge-ready file.
Multiple identity signalsEmail · phone · name · company · address
Every merge is reviewedNothing changes until you approve it
Export the decisionsClean file · merge map · audit receipt
CRM Duplicate Finder & Safe Merge Planner
FIND THE MATCH.
REVIEW THE MERGE.
The matcher combines exact and fuzzy identity signals, but weak evidence such as shared company domains or generic inboxes cannot approve a contact merge. Every group waits for your decision.
Drop in a CRM export
CSV, XLSX, XLS or ODS. Your raw records are parsed and compared locally in the browser. No signup, CRM connection or AI upload.
Matching field setupAuto-detected from headers and values
Review queue0 groups
↯
Select a duplicate group.
Compare the records, choose which one should survive, resolve any conflicting values and then approve the merge or keep them separate.
OutputResolve the review queue first.
The clean CRM exports stay locked while any duplicate group is undecided. The review queue can be downloaded at any time.
How it works
SEE WHY RECORDS MATCH
BEFORE YOU MERGE.
The dangerous part is not finding exact duplicates. It is deciding whether two almost-matching records are genuinely the same entity and what data survives if they are.
Compare identity signals.
Blocking keeps large files practical, then exact and fuzzy evidence is combined. Shared domains and generic inboxes are deliberately weak signals for contacts.
Choose the master record.
Each group shows why it matched. Pick the master record and resolve conflicting field values side by side before approving the merge.
Export the merge plan.
Download the cleaned CRM file, the source-to-master merge map and an audit receipt recording what was merged, rejected and left unchanged.
!
This plans safe data merges. It does not press the irreversible “merge” button inside your live CRM.
HubSpot, Salesforce, Pipedrive and other CRMs have their own object IDs, associations, activity history and merge behaviour. Test the final import or API workflow against your system before applying changes in production.
FAQ
BEFORE YOU MERGE.
A fuzzy match is evidence, not permission. The tool keeps that distinction visible all the way through the export.
Does my CRM export leave the browser?
No. CSV parsing, Excel parsing, identity normalisation, duplicate matching, review decisions and clean-file generation happen in the browser. The tool does not send raw CRM rows to Luna or another AI model.
What counts as a strong duplicate?
Strong contact groups need decisive evidence such as the same non-generic email, phone or record ID, plus no material conflict across the group. A shared company domain, similar company name or common mailbox alone cannot make two contacts a safe merge.
Why can a three-record group be downgraded even when two records match strongly?
Because transitive matching can create dangerous bridges. A group is only labelled strong when every pair in a reasonably sized cluster meets the strong rule. Otherwise the whole group stays in human review.
What happens to conflicting values?
The recommended master keeps its value by default, but every conflict is shown as a selectable field decision. Blank master fields can be safely filled from the only non-blank value in the group. Nothing silently overwrites a populated conflicting field.
Can I keep a suggested duplicate group separate?
Yes. “Keep separate” is a first-class decision and is recorded in the review queue and audit receipt. The clean export then retains every record in that group unchanged.