Shopify product pages built from one approved record
New products launch often, and each one needs a long Shopify record: variants, tax, ingredient list, warnings, shelf life, SEO and metafields. We built an n8n workflow that turns an approved product record into a complete Shopify product, one at a time.
Products wait in a queue and go through one by one. A failure resets one product, not the whole batch.
- Team
- Product and content
- Built with
- n8n
- Shopify
- OpenAI
- Postgres
- Slack
The problem
A new product meant an afternoon of form fields.
The product data already existed in the product database, approved by the team that sources it. Getting it into Shopify still meant copying field by field, writing the product text and SEO, and filling in the metafields the storefront reads.
Products like these carry legal text: ingredient lists and warnings have to be right. Copying by hand is where mistakes creep in.
Launches come in batches. When several products were approved at once, the work piled up on one person.
What we built
One approved record as the source
When a product is ready in the product database, a webhook puts it in a queue. Only sourcing data that has already been approved is used.
AI steps for the text
AI steps in the workflow write and format the parts a shopper reads, such as the product text and the SEO fields, from the approved data.
A complete Shopify product
The workflow creates the product with SKUs, tax settings, ingredient list, warnings, shelf life, SEO fields and metafields, so nobody has to fill in fields by hand afterwards.
One product at a time, with a reset
A poller takes one product from the queue each minute. If a step fails, an error workflow alerts Slack, puts the product back in the queue and leaves a comment on the product record.
Guardrails
What keeps bad data off the shop
A product page is public once it is published. These rules keep half-finished products out.
- A person starts it by approving the product
- Only approved sourcing data is used
- One product at a time, so a failure affects one product
- On error: an alert, a reset, and a comment where the team works
Why we built it this way
A queue instead of a rush
Ten products approved at once would mean ten runs competing for the same API limits. A queue takes them one by one and makes a failure easy to trace.
Fail loudly, where people look
The error comment lands on the product record the team already works in, not only in a log nobody opens.
Approved data in, nothing invented
The workflow builds from data the sourcing team has signed off. It formats and writes around those facts instead of guessing them.
Measured
What the team can see
The queue and its failures are visible without opening n8n.
- Products waiting in the queue
- Failures, with a Slack alert and a comment on the record
- Which products were created in Shopify, and when
More case studies
Other systems our team built.
- Customer supportAn AI support desk where a person still decides what gets sentRead the case →
- FinanceA finance back office that reads its own post and supplier mailRead the case →
- OperationsProduction planning that is done before the team arrivesRead the case →
- E-commerce operationsShop, marketplace and warehouse kept in step without anyone watchingRead the case →
- Leadership and engineeringReports that draft themselves and alarms that fire before things breakRead the case →
Related
Still building product pages field by field?
Tell us where your product data lives. We'll show you how much of the Shopify record it can fill on its own.
Book a 15-min call