I’m using the ‘standard’ prebuilt n8n workflow for Firecrawl but it’s not outputting what I want. I’d like someone to look at it together. I don’t have a massive budget and I’d likely only an hour or so of help.
Michael, this is a clean paid one-hour debug if you already have the workflow and a couple of sample target pages.
I would narrow it in this order: Firecrawl crawl mode/limits, extractor schema, n8n item mapping, then the exact output shape you want back. The useful handoff is not account access; it is the workflow export, 1-2 target URLs, current output, and desired output.
I can tighten that into the right Firecrawl + n8n path and leave you with the minimal node changes instead of a full rebuild. If you want me to take the hour, reply here with the sample and target output, or use the oimrqs ops contact page if you prefer email: oimrqs ops - Web, Automation, Data, APIs and AI Workflow Delivery
welcome to the n8n community @michaelba
Feel free to share it here in the community, and we can help you build it together.
I ran into a similar setup and what saved my sanity was adding a tiny delay plus a quick check step after each Firecrawl call so n8n doesn’t race ahead before the scrape is actually ready. If you need tighter control for the human-in-the-loop part, routing everything through a single Google Sheet tab as the review queue kept things way less messy on my end.
I ran into a similar setup and what saved my sanity was adding a tiny delay plus a quick check step after each Firecrawl call so n8n doesn’t race ahead before the scrape is actually ready. If you need tighter control for the human-in-the-loop part, routing everything through a single Google Sheet tab as the review queue kept things way less messy on my end.
I ran into a similar setup and what saved my sanity was adding a tiny delay plus a quick check step after each Firecrawl call so n8n doesn’t race ahead before the scrape is actually ready. If you need tighter control for the human-in-the-loop part, routing everything through a single Google Sheet tab as the review queue kept things way less messy on my end.
Hi,
I saw your post about the standard Firecrawl n8n workflow not producing the output you need.
This sounds suitable for a short diagnostic session. Before changing the workflow, I would first compare the expected output schema with the actual Firecrawl response, then check the extraction prompt, crawl/scrape mode, and any downstream mapping nodes.
If the issue can be reproduced with a non-sensitive URL and a sanitized workflow export, I can review it without needing access to your production credentials.
If you still need help, please share the expected output example and the current output example. I can then confirm whether a one-hour session is realistic.