Shopify + PIM: sync your e-commerce catalogue effortlessly

Shopify is an excellent e-commerce platform. But for retailers managing thousands of multi-supplier SKUs, its native catalogue management capabilities soon reach their limits. Here’s how a PIM connected to Shopify changes the game.

What Shopify does well — and what it doesn’t

Shopify excels at what it was designed for: online shopfront, shopping basket, payment, order management. It has a functional product catalogue for shops of a reasonable size.

What it doesn’t handle well once you have a few thousand SKUs:

• Multi-supplier imports — no native FTP/SFTP connector, no EAN deduplication, no automatic mapping of supplier columns
• Structured enrichment — Shopify meta fields are powerful but complex to manage manually at scale
• Document compliance — no native management of REACH, RoHS, CE, WEEE
• Multi-channel — if you also sell on Amazon, Fnac or Cdiscount, Shopify is not your central source of truth
• Versioning and audit trail — difficult to know who changed what, when, and why

This is exactly where a PIM takes over: it becomes the single source of truth, and Shopify becomes one of the distribution channels.

PIM → Shopify architecture: how it works

The Pixee PIM Shopify connector is bidirectional. The main flow goes from the PIM to Shopify, with an optional upload of sales and stock data.

1. Supplier import into the PIM — your CSV/Excel/XML files arrive in Pixee PIM via FTP or API. Automatic mapping (AI) standardises the columns.
2. Enrichment and validation — AI-powered enrichment of descriptions, translation, variant management, addition of technical attributes, verification of catalogue completeness
3. Push to Shopify — creation or updating of Shopify products with titles, descriptions, images, prices, stock and custom meta fields
4. Variant management — sizes, colours, packaging: the Shopify variant model is populated from the PIM’s product attributes
5. Stock synchronisation — stock movements are synchronised from the PIM to Shopify in real time or via scheduled batches

Who owns which field: the question to settle before connecting anything

The most common failure of a catalogue synchronisation is not technical: it is a disagreement about field ownership. The PIM overwrites a description that was reworked in Shopify, someone rewrites it in Shopify, the next push overwrites it again. Nobody is wrong — the rule simply was never written down.

The safest approach is to sort every field into one of three categories:

• Owned by the PIM — the PIM writes, Shopify reflects: SKU, EAN, technical attributes, base price, stock, product images. A change made in the Shopify admin will be lost on the next push, and that is intentional.
• Owned by Shopify — the PIM never touches these: hand-built collections, editorial pages, customer reviews, theme settings, redirects.
• Shared, but one-way — the PIM sets the value when the product is created, then never rewrites it. That is the right regime for the handle (the URL segment of the product page) and for copy reworked by a writer.

This grid is decided in one meeting and fits on one page. It saves six months of arguments about who overwrote whose description. Once the ERP and other channels come into play, the same grid extends to every piece of data and every system: it is the starting point of the architecture described in our article on real-time multi-channel synchronisation.

Shopify meta fields: the bridge between PIM and technical SEO

Meta fields are key to displaying structured data on your Shopify product pages — material composition, dimensions, certifications, technical specifications — without hard-coding them into the description.

Pixee PIM maps your PIM attributes directly to the Shopify meta fields you have defined. Once the mapping is configured, all attributes are synchronised automatically with every catalogue update.

The practical benefits:

• Rich product pages without manual entry in Shopify
• Structured data and rich snippets optimised for SEO
• Collection filters powered by structured attributes
• Consistent technical data across all your channels

Variant management: the most complex scenario

Shopify manages variants via product options (size, colour, material). Pixee PIM translates your variable attributes into Shopify options — the general model is covered in our article on managing variants in a PIM — using the following rules:

• Automatic grouping — SKUs sharing the same parent EAN are grouped into a single Shopify product with variants
• Attribute inheritance — common attributes (brand, category, generic description) are at product level; variable attributes (size, colour, weight) are at variant level
• Images per variant — if you have specific images for each colour, they are automatically assigned to the correct variant

Multi-store and multi-currency

If you operate multiple Shopify shops — one per country, per brand or per customer segment — Pixee PIM manages multiple Shopify connectors in parallel from the same central catalogue.

Each connector can have its own rules:

• Different product selection (certain products published in FR but not in DE)
• Store-specific pricing (channel-specific pricing rules)
• Dedicated translations (FR description for the France shop, DE for Germany)
• Independent publication statuses per shop

Stock synchronisation: real-time vs batch

Pixee PIM offers two methods for synchronising stock with Shopify:

• Real-time (webhook) — as soon as a stock movement is recorded in the PIM, Shopify is updated within 30 seconds. Recommended for tight stock levels or fast-moving products.
• Scheduled batch — a synchronisation every X hours (15 mins, 1 hour, 4 hours, 24 hours, depending on your choice). Sufficient for most B2B distributors.

Push deltas, not the whole catalogue

The Shopify Admin API is rate-limited: a shop only absorbs a bounded number of calls per second, whoever is making them. This is not an implementation detail — it dictates how a catalogue gets pushed.

Re-sending every product on every cycle saturates the queue without telling the shop anything new: on an established catalogue, the share of products that actually changed between two synchronisations is small. A properly designed integration therefore sends only the products whose tracked fields changed since the last successful push, and spreads its calls at the rate the API accepts. Three things to check before choosing a connector:

• A large supplier import must not freeze the shop — updates are spread out, not fired in a burst.
• Images are the most expensive item — they should only be re-sent when the source file has changed, never because another field on the product was touched.
• A full resynchronisation is still needed occasionally — after a restore, a mapping change or when taking over an existing shop. It must be triggerable without reconfiguring everything.

Three traps in the first synchronisation

The Shopify shops you connect to a PIM are almost never empty. Three points are settled before the first push, never after.

• The matching key — SKU or EAN? Until it is settled, the first push creates duplicates instead of updating what exists. If the Shopify SKUs were typed by hand for years, the migration starts with an export and a collision check; making the codes reliable is covered in our article on GS1 barcode validation.
• URLs that are already indexed — a Shopify product’s handle is the URL segment of its page. Regenerating it from the PIM title changes the address of pages that already receive traffic. On an existing catalogue, keep the handles in place and only generate them for new products.
• What happens to withdrawn products — “this product is no longer in the PIM” does not mean “delete it from Shopify”. Setting it to draft or archiving it preserves the URL, the reviews and the order history; deleting it does not.

Operational savings: an illustrative scenario

The orders of magnitude below describe a typical situation, reconstructed from setups we commonly encounter. They are not the measured results of a named customer.

Take a Shopify shop with 8,000 active SKUs and 12 suppliers:

• Before PIM: two full-time staff working on updating the Shopify catalogue, about three weeks to integrate a new supplier, numerous data errors (incorrect prices, incorrect stock levels, missing descriptions)
• After PIM: roughly half a full-time equivalent on the catalogue, a new supplier integrated within a few days once the initial mapping is configured, and data errors reduced to a marginal level

On which plan is the Shopify connector available?

The Shopify connector is available from the Starter plan (1 connector included). The Growth (3 connectors) and Scale (up to 6) plans allow you to connect multiple Shopify shops or add other channels (Amazon, Mirakl, Odoo) without additional development.

Frequently asked questions

Should editing products in the Shopify admin be forbidden?

Not necessarily, but everyone needs to know what will be overwritten. Fields the PIM owns revert to their PIM value on the next push: correcting them in Shopify is at best pointless and at worst a repeated waste of time. In practice you leave access open and publish the list of affected fields — a technical constraint is better explained than forbidden.

What happens to products already in Shopify?

They are matched against the PIM catalogue on a unique key before anything is written. Matching produces three sets: products present on both sides, which will be updated; products only in the PIM, which will be created; and products only in Shopify, which need a human decision. The third set is the one that takes time, and rightly so: it holds the shop’s history.

Does the PIM handle SEO on Shopify product pages?

It handles the raw material: structured attributes, descriptions, meta fields the theme can turn into structured data. Title tags and meta descriptions can be fed from the PIM, but that is a decision to make explicitly — a shop whose SEO copy was reworked page by page is better off keeping it under Shopify ownership. The rule does not change: one field, one owner.

Is a PIM justified with a single Shopify shop?

Not always, and the number of shops is not what decides it. What justifies a PIM is the number of inputs — suppliers, file formats, units and reference data that differ — more than the number of output channels. A shop fed by two regular suppliers is perfectly manageable in Shopify alone. The same shop fed by twelve suppliers with heterogeneous formats no longer is, even without a second channel.