I got tired of manually hunting for Instagram trends, so I automated the whole thing in n8n
Every time I wanted to find what was trending on Instagram, I’d end up opening the same accounts, scrolling through Reels, checking views, saving videos, reading captions… and then trying to remember why I saved half of them.
So I built a Reels Trend Watcher in n8n.
Now the workflow can:
→ Pull a list of Instagram accounts from Notion
→ Find their recent Reels with Apify
→ Track views, likes, comments, saves and shares
→ Detect new vs existing Reels
→ Download new videos
→ Send the actual video to Gemini for analysis
→ Extract the hook + full transcription + category + format
→ Translate the content
→ Store everything back in Notion
So instead of just having:
“This Reel got 80K views.”
I can start getting:
“This Reel got 80K views + this hook + this format + this topic.”
That’s where it gets interesting.
Because once you have enough data, you can start asking better questions:
Which hooks keep showing up in high-performing Reels?
Which formats are actually working?
Which topics are suddenly gaining traction?
Are certain accounts consistently outperforming their baseline?
The scraping part was fairly straightforward.
The harder part, IMO, is figuring out what to do with all the data after you collect it.
Right now I’m using n8n + Apify + Gemini + Notion for the workflow.
Next step is turning this from a “Reels database” into an actual trend detection system that tells me what patterns are worth paying attention to.
If you were building this, what would you prioritize first:
1. Viral hook detection
2. Trend/topic detection
3. Competitor monitoring
4. Predicting which Reels might take off