Home · Business · Sales · Lead qualification

YES

As of 13 August 2026, AI can detect duplicate leads in your CRM.

This still needs a person who signs their name to it.

Can you do it?

15 minutesto a draft.

30 minutesto something you’d act on.

Cost, all in£0

Skill neededpower-user

Who has to check ityou

What the alternative costsA CRM administrator can perform the same matching from an export, but the supplied tool data gives no price for that work.

If this goes wrong: you merge two different leads or delete useful history, then sales activity and ownership are recorded against the wrong person.

What to actually do

  1. Hand it to a person

    The route this page recommends

    A person who owns the outcome does this end to end, worth it when the failure is dear.

  2. Use a tool built for this

    Second choice

  3. Do it yourself

    The distant third

    A chat interface, power-user skill, and roughly 30 minutes until you can act on the result.

    How to actually do it

    1. Export your leads from the CRM with a stable record ID and fields such as email address, phone number, name, company, website, job title, location, owner and creation date.
    2. Remove fields the chatbot does not need, and replace unnecessary personal details with record IDs where possible while keeping the values needed to compare records.
    3. Paste the export into a chatbot with the prompt above, splitting it into labelled batches if the full file is too large.
    4. Save the returned groups of definite, likely and possible duplicates in a separate spreadsheet without changing the CRM.
    5. Open each proposed pair in the CRM and compare the full record, including conversation history, owner, consent status and recent activity.
    6. Mark each pair as merge, keep separate or needs review, then merge only the pairs you have confirmed using the CRM's own duplicate-handling process.
    7. Run a second CRM search for the same email addresses, phone numbers and company records to check that confirmed duplicates were handled correctly.

    Prompt

    I need to identify likely duplicate leads in a CRM export. Compare the records below without changing or deleting anything. Use email address and phone number as strong identifiers, then compare name, company, website, job title and location as supporting evidence. Group records into: definite duplicate, likely duplicate, possible duplicate and not a duplicate. For every proposed match, show the record IDs, the matching fields, the conflicting fields and a short reason. Treat shared company details alone as insufficient proof that two people are duplicates. Do not invent or normalise missing facts without showing the original values. Do not recommend merging records when the evidence is weak. Finish with a list of records that need a human decision. Here is the CRM export: [paste CSV or table here].

    Open it prefilled in ChatGPT or Claude, or copy it into Gemini, which takes no prefill link.

What it gets wrong

  • AI cannot know whether two similar records represent the same person when the CRM lacks a reliable identifier.
  • AI cannot decide whether sales history, ownership or consent records should be preserved when two records are merged.
  • AI cannot safely apply bulk merges without risking false matches and lost CRM history.
  • AI cannot resolve conflicting records by asking the lead, account owner or sales team for context.
  • AI cannot guarantee that a CRM export contains every relevant record or reflects the current database.

Even on a YES, the friction has a name: judgement under ambiguity, private data access and verification cost.

How we scored this

Five axes, each scored nought to two by hand: ten means AI carries the task cleanly, and the thresholds that turn a total into YES, PARTLY or NO are published in the methodology. Each axis name links to its definition.

AxisScore (0–2)
Output2
Inputs2
Verification1
Liability1
Effort delta2
Total8 / 10

The methodology and its thresholds are published in full.

FAQ

Can ChatGPT find duplicate contacts in my CRM?
Yes. Export the relevant records and ask it to compare strong identifiers such as email addresses and phone numbers before using names and company details as supporting evidence. It can produce a useful review list, but you should confirm every match in the CRM before merging anything.
How do I find duplicate leads in Salesforce?
Export leads with stable IDs and fields such as email, phone, name, company and website, then use AI to group likely matches without changing the source data. Check the proposed pairs in Salesforce and apply merges only after confirming that the activity, ownership and consent history belong together.
Can AI merge duplicate CRM records?
AI can recommend which records appear to be duplicates, but it should not make unreviewed bulk merges. A false match can combine two different people or lose useful sales history, so the final merge decision stays with you or your CRM administrator.
What fields are needed to detect duplicate leads?
Email address and phone number are the most useful starting points, with name, company, website, job title and location providing supporting evidence. Include a stable CRM record ID so you can trace every suggestion back to the original record.

Nearby answers

All Lead qualification rulings

Assessed by gpt-5.6-luna (gpt-5.6-luna) on 2026-08-13, second-checked by an independent model. Wrong somewhere? Email [email protected] and it gets re-checked.

The newsletter

AI news, new answers and product picks, straight to your inbox.