Map prospect sites into outreach angles with n8n

They built an n8n workflow that collects company URLs, reads prospect pages, and turns them into sales-ready opportunity summaries.

Done with n8nยท Intermediate

What they did
They set up a chat-triggered n8n flow with several smaller agent steps instead of one large prompt. One agent on o4-mini chose likely company URLs, Bright Data's Web Unblocker pulled each page's HTML, and a Code node stripped the page content down for processing. Then separate gpt-5 agents summarized each page, a Merge node brought those outputs together, and a final prompt turned the combined material into business opportunities.
How it went
The workflow returned a set of opportunities for each prospect site that sales teams could use to shape outreach. When they tried to do everything in one prompt, the output was weak and poorly organized.
Tips
Break the job into stages: URL finding, page-by-page summarizing, then final synthesis. Use a cheaper model for the URL-filtering step, and keep the workflow in a test mode before using it on more prospects.

Try it yourself with n8n

Create an n8n workflow that takes [industry terms or search queries], finds likely company website URLs, opens each homepage and about page, and summarizes what each company does plus 2-3 ways [my product or service] could help. After drafting the outreach angle for each company, stop and ask me to approve the results before you save anything, send anything, or submit anything. Start with a test run of [5] companies, then show the output in a table with company name, URL, summary, and outreach angle, and wait for my approval before any further action.

Fill in the [brackets]. Agents can make mistakes, so review anything before it is sent, paid or deleted.

Read the original on dev.to โ†—

Get 3 agent uses like this every morning

With the day's top AI news. Free, and one click to unsubscribe.