Automating your news monitoring with Make and GPT

I used to spend 1 hour reading news every day. To retain 5% of it. Most of it was noise. Today, I receive a 3-minute ultra-targeted briefing every morning on topics that actually matter to my business. Make and GPT do the work.

The Problem with Traditional News Monitoring

RSS, newsletters, Twitter, LinkedIn — every source generates volume. The human brain isn't equipped to filter 200 articles per day and extract the essentials. You end up ignoring everything or reading everything — both are failures.

The System Architecture

• Sources: Targeted RSS feeds + Google Alerts + Twitter/X feeds
• Aggregator: Make (formerly Integromat) — fetches and consolidates news every morning at 6am
• AI filter: GPT-4 analyzes each article, relevance score 1-10, summary in 2 sentences
• Delivery: Only articles scoring ≥7 land in my WhatsApp briefing

The Make Scenarios

Three chained scenarios: (1) Collect — fetches articles from the last 24 hours across all sources. (2) Analyze — sends each title + summary to GPT with a scoring prompt. (3) Synthesize — compiles the retained articles into a formatted briefing and sends it via WhatsApp.

The Scoring Prompt

This is where everything plays out. A good filtering prompt must reflect your precise business context — not a generic "is this interesting" prompt. Mine contains my priority topics, my current projects, and a list of signals to detect (competitor news, regulatory opportunities, emerging trends).

Results After 3 Months

Monitoring time: from 60 minutes to 8 minutes per day. Quality of information retained: significantly higher. Fear of missing something: zero.

The full tutorial with exportable Make scenarios and the scoring prompt will be available soon. Join the list to access it first.