What the "similar usernames" warning means and how to respond to it.
When trackagoat detects that three or more creators in a project share a common username pattern — a shared prefix like wyndywith*, a shared suffix like *_official, or handles that are very close in spelling — it shows a Similar usernames detected warning on the Creator Insights page.
TikTok's recommendation algorithm and API behavior can be affected by network-level account relationships. Accounts that share naming conventions, cross-promote frequently, or appear to operate as a coordinated cluster may be:
Heuristic, not a guarantee
This is a heuristic — not every similar-named cluster is penalized. But if the warning also shows a depressed scrape success rate for the group (below the project average), there is empirical evidence that the clustering is affecting data reliability.
Separate projects: Move related-named creators into their own project. This isolates their data and reduces the risk of cross-cluster throttling during scraping.
Use campaigns for grouped attribution: If you want to track the collective performance of a creator family (e.g. all wyndywith* accounts as a single campaign), create a campaign that includes all of them. Campaign analytics aggregate across members without requiring them to share a project.
Encourage handle diversity: When onboarding a new creator collective, suggest that each creator's handle be distinctive. This avoids triggering TikTok's network-detection heuristics and keeps scraping reliable.
Dismiss the warning: If you've reviewed the cluster and decided it's not a concern, use the × button to dismiss the alert for this project. The dismissal is stored locally per browser.
The detection algorithm groups handles using three strategies:
| Strategy | Detects | Example |
|---|---|---|
| Common prefix | 3+ handles sharing 5+ leading characters | wyndywith* |
| Common suffix | 3+ handles sharing 5+ trailing characters | *_official |
| Edit distance | 3+ handles within 2 character edits of each other | brandchef / brendchef |
Severity is set to warn when the cluster's 48-hour scrape success rate falls more than 20 percentage points below the project average. Otherwise it is shown as an informational notice.
