The problem
A marketing manager spends hours every week manually browsing websites, social pages and news feeds to know what is happening on the market. The information is scattered, the tracking is irregular, and decisions get made on outdated data.
The solution, step by step
- List the 10 to 15 public sources to watch: market news sites, sector RSS feeds, public pricing pages.
- In n8n, schedule a workflow every Sunday night that fetches each source (RSS and HTTP nodes).
- Clean the raw content, then send everything to ChatGPT with a fixed prompt: "Summarize the week's changes on prices, new offers and announcements. Flag anything urgent."
- Ask the model for a structured output: three sections — price moves, new offers, signals to watch.
- Push the brief to a Notion database page dated automatically, so the history builds up week after week.
- Add a condition: if the AI flags something as urgent, send an immediate notification to the team.
- Between 4 and 6 hours saved per week for the marketing team.
- A written, dated market history you can search anytime.
- Faster reactions: pricing or offer changes are spotted within days instead of weeks.