Instagram Reels Outlier Detection & Telegram Reporting

Built an automated Instagram analytics and outlier detection system that monitors 50+ business accounts, analyzes newly published Reels, identifies unusually high-performing content, and sends qualified opportunities directly to Telegram.

The system replaced hours of manual Instagram monitoring with an automated data pipeline running on n8n.

The client managed multiple Instagram business accounts and needed to identify high-performing Reels quickly.

Previously, the team had to manually review content across 50+ accounts, making it difficult to consistently identify emerging viral content and respond while it was still relevant.

The process took 5+ hours of manual work and had no centralized system for tracking previously reviewed Reels.

I designed and built an automated monitoring pipeline using n8n, SocialCrawl API, Google Sheets, Telegram Bot, and JavaScript.

The system establishes a performance baseline for each Instagram account and uses that baseline to identify genuine outliers rather than relying on a fixed global view threshold.

The workflow retrieves the list of Instagram accounts from Google Sheets.

SocialCrawl collects the 12 latest Reels from each account.

The data is normalized and processed inside n8n.

The median view count is calculated for each account.

Each Reel is scored against its account-specific median.

Reels reaching at least 3× the median and published within the last 14 days are identified as outliers.

Previously processed Reel IDs are checked to prevent duplicate notifications.

Qualified Reels are sent to Telegram and recorded in Google Sheets.

The automation uses several decision rules to improve signal quality:

12-Reel baseline: Recent content establishes the account's normal performance.

3× performance threshold: Only significant outliers qualify.

14-day maturity window: Older content is excluded from detection.

Deduplication: Reel IDs prevent repeated notifications.

Rate-limit handling: Requests are batched with controlled delays.

Error handling: Private accounts and API limitations are skipped without stopping the workflow.

Each qualified Reel generates a structured Telegram notification containing:

Instagram handle

Industry category

Publication date

View count

Performance factor

Direct Reel URL

A Google Sheets log maintains the processed Reel data and provides a centralized record of detected content.

Monitors 50+ Instagram business accounts

Processes approximately 600 Reels per run

Saves 10+ hours per week

Identifies approximately 20 qualified Reels per run

Prevents duplicate notifications

Enables the client to identify high-performing content within hours

Allows new accounts to be added directly through Google Sheets

The workflow currently runs in production and is designed to scale by adding or removing Instagram accounts through the central Google Sheets configuration.