Get started for freeGet started
Get started for freeLog in

Duplicate detection

A duplicate spotted
as soon as it arrives.

A supplier resends a reference already in the catalog, a manual entry recreates an existing product. Upsellr spots duplicates as soon as a product arrives, by its identifiers then by its name and brand, and asks you what to do with them.

Get started for freeSee the demo

01The starting point

The same product,
twice in the catalog.

A supplier file resends known references, a manual entry recreates an existing product. The duplicate goes unnoticed: two pages to maintain, two versions drifting apart, and a customer who sees the same product twice.

  • Two pages to maintainEvery correction is made on one and not the other.
  • Versions that drift apartThe same product ends up with two descriptions.
  • A catalog that swellsExtra products after every import.

02How it works

You ask. The agents take care of it.

  1. 01The product arrivesThrough an import, an entry in the grid or a request in the conversation.
  2. 02Checked against the catalogBy its identifiers, such as the EAN or the reference, then by its name and brand.
  3. 03Remove or keepRemove the duplicates or keep them: nothing is merged without you.
  • EAN
  • Reference
  • Name
  • Brand
  • Import

03The result in Upsellr

A duplicate flagged,
a decision to make.

Duplicates are flagged to you in the conversation, with the matching criterion.
You remove them or keep them, in one click.

  • On arrivalImport, entry, conversation
  • ExplainedThe matching criterion
  • DecidedBy you

Illustrative example · sample data

04What makes the difference

Duplicates, spotted before they settle in.

  • Identifiers first.

    EAN, reference: a product that shares an identifier with an existing product is flagged. Failing that, the name and brand take over.

  • Spotted at every arrival.

    After an import, an entry in the grid or a request in the conversation, every new product is checked against the catalog.

  • You decide, nothing is merged.

    The duplicate is presented to you with the matching criterion: you remove it, or keep it if they really are two products.

05Results to document

What you’ll be able to measure.

  • Number of duplicates in the catalog
  • Time spent finding them by hand
  • Pages that drift apart for the same product

Your questions.
Practical answers.

What you need to know before your first request.

How is a duplicate spotted?

By its identifiers first, such as the EAN or the reference, then by its name and brand. A search suggests candidates, and each one is double-checked before being flagged.

Are duplicates merged automatically?

No. You choose to remove the duplicates or keep them. Nothing is ever merged without you.

When does detection happen?

Right after an import, an entry in the grid or a creation requested in the conversation. It doesn’t block the import: it comes afterwards.

What does “Remove duplicates” do after an import?

It deletes the products created by that import that were already in the catalog. Your existing products stay in place.

What about products arriving through a connector?

Detection on store and PIM syncs is coming soon.

One product,
one page.

Try it on your products or explore duplicate detection in a personalized demo.

Import my first fileSee the demo