Features
Powerful tools for complex product data.
Catalog, schema, import, export, shop, assets, completeness, localization and governance — what PIM-TNG actually ships.
Catalog
- Search, filter and sort across the whole catalog.
- A category tree you can reorganize without orphaning products.
- Variants, custom entities and SKU rules for what a flat table cannot hold.
Under the hood
- Products
- One record for every sellable product, with fields organized into tabs you define. List views let you search, filter, sort, and move through the catalog quickly as it grows.
- Categories
- A category tree you can reorganize as needed. Products stay linked to their categories, even when you move an entire branch.
- Taxonomies
- Named field sets you attach to a category. Attaching one decides which fields a product in that branch is expected to carry, so it drives both product creation from taxonomy and the completeness rules, instead of every category being configured by hand.
- Variants
- A variant links to its parent and inherits the fields you choose. Change the parent once, and every variant stays in sync.
- Custom entities
- Suppliers, materials, certifications, care symbols: records with their own fields that products can point at, instead of a free-text column that drifts.
- SKU definitions
- Describe how your codes are built and the catalog can read a SKU back into its parts, or build a new product from one.
- Bulk edit
- Change a field across a selected set of products from the list view, rather than exporting to a spreadsheet and importing the answer back.
- Advanced search and saved views
- Build a query out of conditions combined with AND and OR, then keep the result as a private view in the sidebar so you do not have to rebuild it next time.
- Dashboard
- Catalog KPIs on one screen, with the translation and completeness gaps called out rather than left to be discovered product by product.
Schema and the field engine
- Field types, tabs and units you define yourself.
- Constraints checked on every write, whoever is writing.
- Calculated and virtual fields instead of a copied-down column.
Under the hood
- Field types
- Text, numbers, booleans, dates, single- and multi-select from managed option lists, references to other records, and media.
- Tabs
- Fields are grouped into tabs you name and order, so an editor sees a form built for the job rather than three hundred inputs.
- Dimensions and units
- A measurement carries its unit. Convert on the way out, so the marketplace gets centimetres and the print catalog gets millimetres from the same stored value.
- Constraints
- Required, range, pattern and option-list rules enforced when values are written — by a person, by an import, or by the agent.
- Calculated and virtual fields
- Calculated fields store their result and recompute when inputs change; virtual fields derive on read and store nothing. A failed computation is reported, never silently replaced by the previous value.
- Formulas
- Expressions over the rest of the record: concatenate, convert, round, branch on another field. Written once against the schema, applied to the whole catalog.
Data in
- Upload a workbook and the importer proposes a mapping column by column.
- It asks where it cannot decide, and remembers the answer.
- Dry run first, and the file analyzed is bound to the file imported.
Under the hood
- Guided setup
- Upload a workbook and the importer figures out the structure: which sheet to use, which column identifies the product, and what the rest of the data represents. If anything is unclear, it asks you to choose.
- Column mapping
- Map to existing fields, create new ones on a tab of your choosing, or skip the whole column. New fields are proposed with a type and a home before a single row is written.
- Shapes beyond one row per product
- Attribute-per-row (EAV) files, asset sheets, image filename columns and custom-entity lists are all recognized shapes, not special cases you have to reformat by hand.
- Matrixify Shopify imports
- A Matrixify export interleaves parents and their variants down a single column. The importer recognizes that shape and normalizes it into products and variants, instead of asking you to split the workbook first.
- Category paths
- A cell reading "Tools > Drills > Cordless" creates the whole chain and links the product to the leaf.
- Dry run and provenance
- Every import previews what it would create, update and skip. The file analyzed is bound to the file imported, so a mapping cannot be applied to a different spreadsheet by accident.
- Saved mappings and schedules
- Save a mapping for each supplier and reuse it automatically on the next file. Add a schedule, and the drop folder becomes a repeatable integration.
Data out
- A channel is a template plus per-column sources, transforms and conditions.
- Export everything once, then only what changed.
- XLSX or CSV, a live push into Shopify Admin and PDF, on a schedule if you want one.
Under the hood
- Channels
- One per destination, each with its own column list, its own names and its own order. Import their template and map against the real headers.
- Transforms
- Per-column sources, transforms and conditions: pick a field, convert a unit, join several, fall back when empty, emit nothing unless a condition holds.
- Delta exports
- After the first full run, export only the products that changed. The catalog tracks what each channel has already seen.
- Formats
- XLSX or CSV to download, a push straight into Shopify Admin, and PDF written to the asset library.
- Shopify Admin push
- Publish into a store over the Admin GraphQL API using your Admin token, so the channel ends in the store rather than in a store-shaped file somebody still has to upload.
- Scheduled runs
- Recurrent exports run on their own and leave a record of what went out and when.
- Public media URLs
- Exports can carry stable public links to images and documents, so a channel gets a URL that keeps working.
Shop Beta
You already have the product data, so why not sell from your own shop.
- A storefront that sells from the live PIM catalog — no second product database.
- Cart, shopping lists, quick order and checkout for guest or invited buyers.
- Editable themes.
Under the hood
- Catalog without a copy
- The shop reads products over a catalog API. Publish allowlists and price/stock binds decide what buyers see.
- Buyers
- Give your customers access to the shop. And decide if walk-ins can order.
- Lists and quick order
- Named shopping lists beside the checkout cart. Paste SKU and quantity rows to fill a cart or a list in one go.
- Checkout and orders
- Customizable shipping and tax according to your needs. Payment and fulfillment status are tracked separately.
- Payments
- Bank transfer, payment terms, and payment processors at checkout.
- Custom Domains
- Link your PIM-TNG shop to your own domain.
Assets
- Register media by URL, or by filename and attach the bytes later.
- Renditions, versions and rollback in a click.
- Every image knows its product, its role and its position.
Under the hood
- Ingest by URL or filename
- Register media from a link, or import a column of bare filenames as placeholders and attach the bytes later.
- Renditions
- Derived sizes generated for you, with warnings when a stale one would be reused instead of rebuilt.
- Versions and rollback
- Replacing an image keeps the previous one. Rolling back is a click, not a restore from backup.
- Roles and order
- Per product, an asset is the main image, a gallery shot at a given position, or a document. Channels export the role they asked for.
Completeness
- Completeness scored per product against what the catalog requires.
- Constraint failures appear immediately, wherever the change comes from.
- Anything that needs human review becomes a task you can assign.
Under the hood
- Completeness
- Scored per product against what the catalog requires, and kept honest as the schema moves — add a required field and the numbers update rather than quietly staying green.
- Validation
- Constraint failures appear immediately, whether the change comes from an editor, an import, or the agent.
- Tasks
- Work that needs a person — a missing photo, an untranslated description — becomes an assignable item instead of a note in a chat thread.
Localization
- Multilingual interface.
- Translate fields to any language.
- Labels, option names, numbers and dates follow the locale.
Under the hood
- Interface languages
- The application is available in all European languages, with each user choosing their preferred language.
- Data locales
- Translatable fields hold a value per locale, so a description exists in each market's language without duplicating the product.
- Translated schema
- Field labels, tab names and option names carry their own translations — the editor reads them in the user's language, whatever language the data itself is in.
- Formats that follow the locale
- Numbers, dates and units are rendered the way each market reads them, in the interface and on the way out.
Governance
- Access by role, with administration behind its own tiers.
- Every agent run and every import kept on the record.
- Deleted products leave a tombstone for downstream channels.
Under the hood
- Roles and tiers
- Access by role, with administration and platform settings behind their own tiers rather than a single all-powerful account.
- Agent activity
- Every agent run is recorded — the tools it called, the arguments, the outcome — and readable afterwards, run by run.
- Import history
- Each import keeps its file, its mapping and its result, so "where did this value come from" has an answer.
- Deletion records
- Deleted products leave a tombstone, so downstream channels can be told a product is gone rather than just finding it missing.
- REST API
- Everything the interface does is available under
/api/v1, authenticated with a bearer token that carries exactly its owner's permissions — the same ones the UI would apply. The OpenAPI description is served at/docs. - MCP
- The same tool set is exposed over MCP, so an agent running outside the app — in Cursor or the Claude CLI — works with the catalog through the same permissions and leaves the same record as the built-in one.
- Cold backups
- Cold copies of the database and the media originals go offsite, with the retention period shown on the pricing page. The security page has the detail.
AI
- This section is so important that we created a full page for it.
Want to see it against your own data?
The fastest way to judge a PIM is to watch your own worst spreadsheet go through it.