Outlier Detection
Outlier detection means finding Reels that beat their own creator's usual views by a wide margin, instead of the posts with the biggest raw numbers. The guides below show how to calculate a baseline, spot outliers on competitor accounts, and read the hook, topic and save patterns that tend to sit behind them.
Curated by Shivank Goura, co-founder of Octupie · Last updated October 16, 2026
What to know about Outlier Detection
- Use the median of an account's last 30 reels as its baseline, since one breakout post badly distorts the mean.
- A 3x multiple over that median is a practical working threshold, and applying it consistently matters more than the exact number.
- One outlier is a data point, but three outliers from different accounts sharing a pattern is a repeatable format worth testing.
- Topic windows usually close within a few weeks, and the first two or three creators to publish on a topic tend to capture the outlier reach.
3 guides on outlier detection
- October 16, 2026·Tool Comparisons
Best Instagram Reels Analytics Tools in 2026 (10 Compared)
Ten Instagram Reels analytics tools compared: Insights, Metricool, Socialinsider and Sprout Social for your own numbers, plus outlier tools for your niche.
- May 12, 2026·Content Research
How to Track Competitor Instagram Accounts
Find the right competitor Instagram accounts, spot outlier posts, decode why they worked, and ship your own version before the window closes.
- May 12, 2026·Content Research
What Drives Instagram Outliers in Three Niches: Patterns We've Noticed
Patterns we've noticed on Instagram outlier reels across AI/tech, education, and lifestyle: hook archetypes by niche, save-to-like signal, topic windows.
Common questions.
01What is an outlier post on Instagram?
It's a Reel that significantly outperformed that specific account's typical reach, not the most-viewed post on the platform. A 200k-view Reel on a 20k-follower account is an outlier, while the same views on a 2M-follower account is below average. The signal is the ratio of performance to the account's own baseline, which takes follower count out of the picture.
02How do I calculate my Instagram baseline?
Sort your last 30 Reels by view count and take the median, which is the 15th post in that list. Use the median rather than the mean: if 29 Reels get 5,000 views and one gets 500,000, the mean jumps to 21,667, which doesn't reflect what your audience normally sees from you. Judge every new post against the median.
03Does posting more often produce more outliers?
Not on its own. In the accounts Octupie users track, daily posters don't produce outliers at a noticeably higher rate than accounts posting three times a week. Topic timing seems to matter more, meaning posting on a subject while it's generating strong signals on similar accounts. Three researched, in-window posts a week tend to beat daily posts made without research.
