In short
Analyzed videos from a followed account can reveal recurring concepts. Inspect cluster members to verify the repeated structure, then review available volumes and performance. How often an idea is used and how much it outperforms are separate signals to examine.
Inspect repetitions
Open Competitors, choose the account, then inspect available concepts. A repeated concept requires at least three independent videos associated with the same cluster.
Collection alone is insufficient: videos need a usable analysis. Wait for processing before interpreting an absence of concepts.
First check that available posts belong to the correct account and network. Then open cluster members: repetition should concern a recognized structure in several videos, not merely an identical product name or a reused thumbnail.
Understand the best repeating concept
The best repeating concept uses the median of comparable account-relative ratios, with at least three ratios and 60% coverage. The median must exceed 1. The most frequent concept is therefore not automatically the best.
Coverage indicates the share of variations with a calculable ratio. It prevents declaring a cluster best based only on the one video with complete data. If no cluster meets the conditions, no measurable best concept is expected even while repeated concepts remain listed.
Example: frequency and strength of evidence
Fictional example: a cluster contains five variations and three comparable ratios of 2×, 3× and 4×. Coverage is 60% and the median is 3×. It can meet the conditions for best repeating concept. A more frequent cluster with a 1.2× median does not automatically rank first because it appears more often.
After reading that signal, compare examples with usable analyses and metrics. Look for shared elements and production differences. The calculation selects evidence to study; it does not identify a recipe with guaranteed success for your brand.
Check the examples
Open several cluster videos and compare what actually repeats. Repetition indicates observed use; it does not prove the creator made sales or that a new adaptation will achieve the same result.