From a pasted link to a published listing in under a minute
How we replaced a 25-30 minute manual listing process with a single-webhook automation pipeline, and cut cost per listing by 95%.
Every marketplace seller knows the drill: copy the product details, resize and re-upload the photos, write the listing, publish it, then post about it everywhere else. Multiply that by dozens of listings a week and it stops being a task: it becomes a job on its own.
The 30-minute problem
Our client, an Australian marketplace seller, was spending 25 to 30 minutes per listing across half a dozen tools before anything went live. None of it was hard. All of it was repetitive.
- Copying product data from multiple sites by hand
- Downloading, renaming, and re-uploading images
- Publishing to WordPress, then posting to social
- Sending marketing messages and updating a spreadsheet
What we built
Instead of automating each step in isolation, we replaced the whole process with one interface: paste a product link, click start, and walk away. Everything downstream (scraping, image handling, publishing, distribution) runs on its own.
- A lightweight custom interface posts the link to an n8n webhook
- A cloud function scrapes the listing; Python handles the images
- An AI step structures the raw content into clean fields
- The WordPress REST API publishes the listing, and the master sheet updates automatically
The best automation isn't the one with the most steps. It's the one the client never has to think about again.
The stack
| Layer | Tool |
|---|---|
| Orchestration | n8n |
| Scraping & imaging | Python on Google Cloud Platform |
| Publishing | WordPress REST API |
| Source of truth | Google Sheets |
The result
What used to take 25–30 minutes across multiple tools now runs from a single pasted link. The client's team reassigned that time to sourcing new inventory instead of data entry, and cost per listing dropped by 95%.
POST /webhook/new-listing
{
"url": "https://marketplace.example/listing/1284"
}If a process in your business looks like this: repetitive, multi-tool, and bottlenecked by a person moving data from one place to another, it's very likely automatable. That's usually where we start.