Back to resources

AI onboarding

How Magic Catalog Import Works (and Why Your JOOR Export Is Enough)

Bring a line sheet PDF, a CSV, a Shopify catalog, or the export from your current platform. The importer builds the room; you approve the diff.

Onboarding is the moat incumbents rely on

The reason brands stay on platforms they dislike is rarely love — it is the weeks of catalog work they already sank into them. Magic catalog import attacks exactly that: the export file you already have is the input, and a buyer-ready wholesale room is the output.

What the importer actually reads

Line sheet PDFs, CSVs, Shopify catalogs, JOOR and NuORDER exports, and even product photos. The model extracts styles, variants, size runs, colorways, wholesale and suggested-retail prices, and materials — and normalizes them into the marketplace's structured catalog schema.

The diff is the product

Nothing publishes on import. Every run produces a field-level review diff: what was read, what was inferred, what the importer is confident about, and what it held for your review — a missing fabric weight, an ambiguous HS code. You approve the diff; only then does the catalog change.

Enrichment follows, on the same terms

After import, AI listing enrichment drafts titles, descriptions, and search tags with cited evidence and confidence bands, queued for your approval. The pattern is constant across the platform: the AI assembles, a human decides.

Minutes, then verification

From export upload to a reviewable room is measured in minutes, not weeks. From there the readiness score tells you exactly what is missing before buyers see the room — and the free wholesale readiness audit gives you the same scoring before you even create an account.

Next step

Turn the idea into the right onboarding path.

Use the public role paths to move into buyer onboarding, brand onboarding, or an operator conversation before gated workspace access.