Automate RSS Feed to MongoDB with Webhook Integration

The Day Maya Hit Her Breaking Point

Maya stared at the spreadsheet on her screen, eyes blurring over rows of links and headlines. As a content marketer for a fast-growing real estate and restaurant SaaS startup, her job was to track industry news, spot relevant articles, and feed them into the company’s internal dashboard.

Every morning started the same way. Open a handful of RSS feeds. Skim through dozens of articles. Copy the ones that mentioned realtors, real estate, or anything about restaurants. Check if the link was already in MongoDB. If not, paste it in. Then send a notification to the team via a custom webhook endpoint so the dashboard could refresh.

It was repetitive, fragile, and painfully easy to mess up. Some days she missed articles. Other days she accidentally added duplicates. And when she took a day off, the whole system fell apart.

That morning, after catching a third duplicate article in the database, she finally said out loud, “There has to be a better way.”

Discovering n8n and a Ready-Made Template

Maya had heard developers around her talk about n8n, a workflow automation tool that could connect services, process data, and run on a schedule. She had always assumed it would be too technical, but frustration is a strong motivator. So she opened her browser and searched for something that could automate RSS feeds into MongoDB.

That search led her to an n8n workflow template: an automation that could read an RSS feed, filter content by keywords, prevent duplicates in MongoDB, and send notifications via a Webhook.

It sounded like everything she had been doing manually, packaged into a single repeatable workflow.

Setting the Stage: What the Workflow Actually Does

Before she imported the template, Maya wanted to understand the big picture. The description broke it down clearly:

• Fetch RSS feed items automatically on a schedule or on demand
• Filter articles by specific keywords, like “realtors”, “real estate”, or “restaurant(s)”
• Process each article one by one, so checks and inserts are reliable
• Check MongoDB to see if an article already exists based on its link
• Insert only unique articles into a MongoDB collection
• Send a POST request via Webhook whenever a new article is stored

It was exactly her current workflow, just automated and much more reliable.

The First Run: Triggers, Feeds, and Filters

Maya imported the template into her n8n instance and began to trace the path of an article through the workflow.

1. Cron & Manual Trigger – The Workflow’s Starting Point

At the top of the canvas, she saw two familiar-looking nodes: Cron and Manual Trigger.

The Cron node was configured to run every hour, which meant n8n would automatically start the workflow without her lifting a finger. The Manual Trigger was there for those times when she wanted to run it instantly, like during a big product launch or industry event.

“So this replaces my morning routine,” she thought. “No more opening feeds manually.”

2. RSS Feed Read – Pulling in the Latest Articles

The next node was RSS Feed Read. It was already set to use the URL:

https://www.feedforall.com/sample.xml

In the template, this feed was just an example, but the behavior was exactly what she needed. The node fetched the latest articles from the feed and passed them along to the rest of the workflow.

“So instead of me scrolling through a feed reader,” Maya realized, “this node is doing it for me, every hour.”

Rising Action: Teaching the Workflow What Matters

Of course, not every article was relevant. Maya only cared about pieces related to realtors, real estate, or restaurants. That was where the IF nodes came into play.

3. Conditional Checks – Filtering by Keywords

The template used a set of IF nodes to inspect each article’s title. Under the hood, these nodes used regex matching to find specific keywords:

• Titles containing “realtors” or “real estate”
• Titles mentioning “restaurant” or “restaurants”

Articles that matched the first condition flowed down one branch. Those that did not match were passed to the next IF node, which checked for restaurant-related terms.

Anything that failed both checks was quietly sent toward an End node, effectively filtered out of her process.

Maya smiled. “So the workflow is literally reading headlines and deciding if I would care.”

4. SplitInBatches – One Article at a Time

Next, she noticed the SplitInBatches nodes. These nodes were responsible for taking the filtered articles and handling them one by one.

Instead of trying to process an entire feed at once, the workflow split the items into batches and processed each article individually. That made it easier to check for duplicates, insert into MongoDB, and send notifications without confusion.

It also meant the workflow could scale to larger feeds without becoming unwieldy.

The Turning Point: MongoDB & Webhook in Action

The real test of the template came when Maya followed the path into the database layer. This was where her manual process had always been the most fragile and time-consuming.

5. MongoDB Find – Avoiding Duplicate Articles

For each article, the workflow used a MongoDB Find node. Its job was to look in her MongoDB collection and see if an article with the same link already existed.

The check used the article’s link property, matching it via regex. If a match was found, the article was considered a duplicate and did not need to be inserted again.

If no match was found, the article continued forward as a candidate for insertion.

6. Merge Nodes – Keeping Only the New Content

To coordinate this logic, the template relied on Merge nodes. These nodes combined the outputs of previous steps so that only genuinely new articles – those not found in the MongoDB search – proceeded to the insertion stage.

In Maya’s old world, this was where she would manually scan the database or trust her memory. Now, the workflow did it deterministically, every hour, without mistakes.

7. MongoDB Insert – Storing New Articles

Once an article passed the duplicate check, it flowed into a MongoDB Insert node. Here, the workflow inserted the article into the articles collection.

Each unique item was stored with its link and other relevant fields, ready to be used by internal tools, dashboards, or reporting systems.

8. Webhook – Notifying the Rest of the System

The final step for each new article was a Webhook node. This node sent a POST request to a specified URL with the article’s link in a JSON payload.

That meant that as soon as a new article was inserted into MongoDB, the rest of the system knew about it. Dashboards could refresh. Alerts could be triggered. Other automations could pick up the new content and run with it.

Articles that were not new, or that failed the keyword checks, were routed to an End node. Their journey stopped there, cleanly and quietly.

The Resolution: From Manual Chaos to Automated Flow

After watching the workflow run once, Maya checked her MongoDB collection. There they were: only the relevant articles, no duplicates, each one neatly inserted. Her webhook logs showed successful POST requests for every new link.

What used to take her an hour of manual work every day was now handled automatically, every single hour, whether she was at her desk or not.

Benefits Maya Saw Immediately

• Automated content ingestion: The RSS feed was read on a schedule via the Cron node, with the option for manual execution when needed.
• Smart filtering: IF nodes with regex checks ensured only articles about realtors, real estate, or restaurants made it through.
• No more duplicates: MongoDB Find and conditional logic prevented repeated entries based on the article link.
• Scalable processing: SplitInBatches handled large feeds by processing items one at a time.
• Real-time notifications: Webhook POST requests alerted downstream systems as soon as new content was stored.
• Flexible execution: Cron scheduling covered routine runs, while the Manual Trigger gave her on-demand control.

How She Tailored the Template to Her Own Needs

Once the base workflow was running, Maya realized how easy it was to adapt it to different projects and feeds.

• Changing the RSS source: She swapped the sample URL in the RSS Feed Read node with her own industry feeds.
• Adjusting keyword filters: She updated the regex patterns in the IF nodes to include additional terms her team cared about.
• Tuning batch sizes: For heavier feeds, she modified the SplitInBatches configuration to control performance.
• Customizing MongoDB: She pointed the workflow to different MongoDB collections and added extra fields for richer data.
• Routing to different webhooks: By changing the Webhook URL, she could notify different services or environments.

What started as a single use case for real estate and restaurant news quickly became a reusable pattern for any type of content her team needed.

Bringing It All Together

Maya’s story is not unique. Many marketers, founders, and developers wrestle with the same problem: keeping data fresh, relevant, and deduplicated without drowning in manual work.

This n8n workflow template offers a clear path out of that loop. It automates fetching RSS feeds, filters content by the keywords that matter, checks MongoDB for duplicates, inserts only new articles, and sends webhook notifications for real-time updates.

Instead of chasing feeds and spreadsheets, you can let automation handle the heavy lifting and focus on what to do with the information, not how to collect it.

Ready to turn your own manual RSS process into a reliable automation?
Set up this template in your n8n instance and start simplifying your RSS feed processing today.