Bulk Changes with Matrixify

Changing many products at once without losing anything — and the backups that make it safe.

The rule that prevents almost everything going wrong

Export first, edit that file, import it back. Never build an import from scratch. The export carries the exact column names, the exact identifiers, and a copy of the data as it was before you touched it.

Matrixify does far more than this guide covers. Their own documentation is thorough and worth reading before anything unusual: https://matrixify.app/documentation/

Back up before you change anything

Two different backups matter, and they are not the same thing.

BackupWhenWhy
The narrow export of what you are about to changeEvery single time, before editingRestores the exact fields you touched
A full product export, all columnsBefore any large or unfamiliar jobRestores anything you did not expect to touch

For a full backup: Apps → Matrixify → Export → Products, all columns, Metafields enabled. Save the file somewhere it will still be there in six months.

The method

  1. Export what you intend to changeApps → Matrixify → Export

    Filter to just those products. Include the columns you plan to edit and enable Metafields if you are changing any.

    Export downloaded

  2. Save an untouched copyYour computer

    Duplicate it and name the copy clearly, e.g. before-2026-08-26.xlsx. Edit the other one.

    An untouched copy exists

  3. Delete the columns you are not changingThe working file

    Matrixify only updates columns present in the file. Keeping just the identifier and the columns you are editing makes the import far harder to get wrong.

    Only identifier plus edited columns remain

  4. Make the editsThe working file

    Edit values in place. Do not reorder or rename columns, and do not touch ID or Handle.

    Identifier columns untouched

  5. Import five rows firstApps → Matrixify → Import

    Cut it to five products, import, and check those five on the live site. Most import mistakes are systematic — five rows show the problem as clearly as five hundred.

    All five correct on the live site

  6. Import the restApps → Matrixify → Import

    Matrixify shows a summary before it commits. Read it and check the row count matches what you expect.

    Row count matches, no failures

  7. Read the results fileMatrixify → Import history

    It lists every row and what happened to it. Failures are reported there, not in the admin.

    No unexpected failures

The columns

Checked against a real Matrixify export rather than assumed. Products live on a sheet named Products.

PurposeColumn
Which product this row isID — Shopify's own number, most precise. Comes from the export
Which product this row isHandle — the web address stub, readable and reliable
Which product this row isVariant SKU — useful when working from the master file
What to do with the rowCommand — usually UPDATE. Also NEW, REPLACE, MERGE, DELETE
Title and descriptionTitle · Body HTML
PlacementType · Vendor · Tags
How tags are appliedTags CommandMERGE or REPLACE. See the warning below
PublishingStatus · Published · Published Scope
IdentifiersVariant SKU · Variant Barcode
ImagesImage Src · Image Position · Image Alt Text · Image Command
Any Jessar fieldMetafield: jessar.<key> [<type>], e.g. Metafield: jessar.romance [multi_line_text_field]

Starter files

Five starter files ship alongside these guides, in a folder named matrixify-starter. Their metafield columns were generated from this store's own field definitions, so the namespace, key and type in every header are correct.

FileFor
jessar-specs-update.csvFilling in specification fields across a category
jessar-copy-update.csvTitle, description, romance copy, bullets and meta beam
jessar-retag.csvAdding shelf tags — pre-set to MERGE so nothing is wiped
jessar-publish.csvChanging status and publication in a batch
jessar-images.csvAdding or replacing product images

If a fresh export ever disagrees with a starter file's column name, the export is right — Shopify and Matrixify both change over time. Tell RevShift so the starter file can be corrected.

When something goes wrong

  1. Stop importingMatrixify

    Do not try to correct a bad import with another import until you know exactly what the first one did.

  2. Read the results fileMatrixify → Import history

    Establish how many products were actually changed, and how.

  3. Re-import the untouched copyMatrixify → Import

    The file you saved before editing restores the previous values for the columns it contains.

  4. Tell RevShift if web addresses were affectedOutside Shopify

    Restoring data does not restore links. If handles or statuses changed, redirects may be needed.