A PIM is a process, not a purchase
The software is the smallest part. What decides success is who enriches what, in which order, and when a product is allowed to go live. We design that workflow with your team, model the attributes per category and market, then connect it to Shopify and your feeds so the same data serves the storefront, the marketplaces and the ads.
PROVEN
RESULTS
TOTAL SALES VS. PREVIOUS PERIOD
Who owns which product data
Most integration problems start with two systems that both believe they own the same field. Before connecting anything, we decide per attribute where it is created, where it is enriched and where it is only read. Typically the ERP owns article numbers, prices and stock, the PIM owns descriptions, specifications, images and translations, and Shopify owns merchandising: collections, sort order and content that only exists in the storefront.
ERP: article numbers, cost and sales prices, stock levels and logistics data
PIM: titles, descriptions, specifications, images, translations and channel-specific copy
Shopify: collections, product status, sales channels and storefront merchandising
Metafields: structured attributes such as material, care instructions or dimensions, typed and validated
Metaobjects: shared entries like size guides, ingredients or designers, maintained once and referenced by many products
Feeds: derived from the same data, never edited separately in a feed tool
The PIM connection and the ERP integration are designed as one data flow. That way a price update never overwrites a translated description, a new product only reaches Shopify once its required attributes are complete, and a sync error ends up in a log someone checks rather than on a live product page.
Choosing a PIM for Shopify
There is no best PIM, only one that fits your catalogue and your team. Akeneo, Salsify and Plytix are common examples. They differ in workflows and approvals, how they manage images and other assets, data modelling, pricing model and the maturity of their Shopify connector. A brand with a few thousand products and one content editor needs something different from a manufacturer with technical specifications in six languages.
We compare options on your own data, not on a feature list. How is a representative product modelled? How do variants map to Shopify options? How are translations and markets exported? And what happens when the connector runs into an API limit or a validation error halfway through a large update?
Sometimes the honest answer is that you do not need a PIM yet. Well-defined metafield definitions, a clear enrichment process and a spreadsheet import can carry a smaller catalogue for years. We say so when that is the case, and set up the metafields so a PIM can take over later without remodelling everything.
Markets, languages and channel feeds
Selling in several countries multiplies product data. With Shopify Markets you can vary language, currency, pricing and product availability per market, but the translated and localised content still has to come from somewhere. We connect the PIM so translations land on the right products, metafields and metaobjects in Shopify, and a missing translation is caught before a market goes live.
Translations per language, including content in metafields and metaobjects
Product availability and assortment rules per market
Google Merchant Center and Meta catalogue feeds from the same source
Marketplace exports with their own category and attribute requirements
Catalogues for Shopify B2B with their own assortment per company
Shopify’s standard product taxonomy and category attributes where they add value
Each channel receives the version it needs from one enriched record, mapped to that channel’s required fields. When Google or a marketplace rejects a product, the correction happens in the PIM rather than in a feed tool. That way the fix also reaches the storefront, the other markets and every future export.
How a PIM integration runs
Product data audit. We analyse a sample of products across categories: which attributes exist, where they live, how complete they are and which fields each channel and market actually requires.
Attribute model per category. We define families, attributes, validation rules and required fields per category and market, and decide which attributes become Shopify metafields, metaobjects or variant options.
Ownership and workflow. Per attribute we record the leading system and the person responsible. We design the enrichment steps and the completeness rules a product must meet before it can go live.
PIM selection or setup. Where no PIM is in place, we test the shortlisted options on real products. Where one exists, we review its configuration against the new model before building on it.
Connection and migration. We configure or build the connector, migrate and clean existing data, and test updates, deletions, translations and error handling with representative products, not only the happy path.
Go-live and governance. We switch channels over in a fixed order, monitor the sync logs and train editors. Afterwards we review completeness and errors periodically, so the model keeps pace with the catalogue.
Shopify product data in numbers (October 2026)
These are the Shopify limits we design a PIM connection around. They decide which data model works, how the sync is built and where Shopify needs a workaround. Figures as documented by Shopify in October 2026.
Options and variants: up to 3 options and 2,048 variants per product, on every plan. A PIM model with a fourth variant axis, such as length next to size, colour and material, has to be split into separate products or move that axis into metafields.
Combined listings: on Shopify Plus and enterprise plans you can group up to 60 products into one listing, with up to 2,048 variants across them. Each colour can then be its own product with its own content and URL, which is how many PIMs already model it.
Media: up to 250 images, videos or 3D models per product, with images up to 25 megapixels. The connector decides which assets go to Shopify and in which size; a full asset library rarely belongs on the product.
Metafield definitions: up to 256 per resource type for the merchant, plus 256 per installed app. Most value types hold up to 64 KB, JSON up to 128 KB, and list fields up to 128 items. Only attributes that the storefront, filters or channels actually use should become a metafield.
Metaobjects: 128 definitions on Basic, Shopify and Advanced, 256 on Plus and enterprise, each with up to 40 fields and 1,000,000 entries. Shared data such as materials or size charts is synced once as an entry and referenced from products, with up to 1,024 references per list field.
Standard Product Taxonomy: the product category unlocks category metafields with fixed entries, such as colour, size or fabric. Mapping PIM values to those entries keeps filters and colour options consistent, instead of free text per product.
Languages: a shop can have up to 20 enabled languages. Translations are written per language with a separate API call (translationsRegister), not with the product update, so the connector has to track which language versions changed.
API rate limits: the Admin GraphQL API restores 100 points per second on standard plans, 200 on Advanced, 1,000 on Plus and 2,000 on enterprise; a single query may cost at most 1,000 points. A full catalogue sync on a smaller plan has to be queued and spread out.
productSet and bulk operations: productSet writes a whole product in one call and deletes variants, metafields and collections that are missing from its input. Bulk imports take a JSONL file of up to 100 MB and must finish within 24 hours; from API version 2026-01 an app can run five at once per shop. Once a store has 500,000 variants, at most 10,000 new variants can be created per day, which sets the pace of a large initial load during a migration.
Collections: up to 5,000 smart collections with up to 60 conditions each, and up to 128 metafield definitions can serve as a collection condition. Assortment rules based on PIM attributes work best through a few well-chosen metafields, not one collection per attribute value.
When you do not need a PIM: with one or two languages, at most a few thousand products, attributes that fit in a few dozen metafield definitions and one person maintaining the data, Shopify is enough on its own. Metafield definitions, metaobjects and a spreadsheet or bulk import app cover the content, and an ERP integration supplies prices and stock. Revisit that choice when you add markets or integrations; the limits above show where Shopify alone starts to pinch.
Product data and integrations in practice
Let’s talk about your product data
Tell us where your product data lives today and which channels and markets it feeds. We reply with a first view on the right PIM setup and how it connects to Shopify.
