Creator Retention & Risk Radar: assess and follow up
Compare monthly performance, review the reasons behind risk scores and record a shared contact and follow-up plan.
This guide describes the available features. Your account role, assigned modules, signed contract and network plan determine which controls you can use. Reading a guide does not change your access.
Start here
- Ensure the monthly leaderboard contains at least the analysis month and its previous month.
- Select Analysis month and choose Refresh.
- Filter Needs attention queue by Risk, Manager or Action status, then select Apply.
- Choose Review on a creator and read the score factors alongside the two months’ metrics.
- Set Action status, Follow-up date and a concise Team note, then choose Save plan.
- As owner, configure Significant-drop alerts only after reviewing the destination and thresholds.
What the score measures
The radar compares the selected month with the previous month using diamonds, LIVE hours, valid days and low/no-activity signals. Critical is 70–100, High 50–69, Medium 30–49 and Stable 0–29. A creator without prior-month data is labelled New / no baseline, rather than being treated as a proven decline.
The score is a reporting indicator. Review the underlying data and the creator’s circumstances before taking action. A partially imported current month can naturally look lower than a complete previous month; the displayed comparison is not a promise that the data periods contain equal elapsed activity.
Owners see their network; sub-users are limited to creators in their allowed hierarchy. Managers can update visible creators’ attention plans. Only the owner controls alert destinations, thresholds and manual alert sending.
Read the score factors
Review expands the row to explain what contributed to the score. Diamond decline can contribute up to 40 points, LIVE-hour decline up to 25, valid-day decline up to 20, and inactivity up to 15. Growth or a missing positive baseline is not treated as a negative percentage decline.
Needs attention counts unresolved scores of 30 or higher. Critical risk and other overview cards provide context, while the queue lets you act on individual records. A creator marked Resolved can still have low imported performance; workflow status and calculated risk are separate.
Coordinate a follow-up plan
- Apply the filters you need.
- Open Review and verify the score explanation.
- Contact or investigate through the appropriate team process.
- Update status, date and note, then Save plan.
- Return on the follow-up date and reassess fresh performance and the agreed action.
- Risk filter
- Choose the level you want to investigate, including stable or baseline categories where offered.
- Manager filter
- Narrow the queue to a manager within the available report.
- Action status
- Open means work is outstanding; Contacted records outreach; Watching keeps the creator under review; Resolved records a completed follow-up decision.
- Follow-up date
- The planned next review date. Use it to avoid leaving a contacted creator without a next step.
- Team note
- Up to 2,000 characters of contact context or the next action. Saving replaces the current attention note; write a useful current summary.
Export and compare retention patterns
Export CSV downloads the queue data using the selected report filters. Check month, manager, risk and action-status filters before sharing the file internally.
Manager retention summarises creator retention for available manager data. Compared with creators who left compares current at-risk averages with historical leavers’ final recorded months. No historical exit cohort means there is not yet enough relevant exit history; it is an empty benchmark, not a zero-risk conclusion.
Configure Significant-drop alerts as owner
Enable alerts and provide a valid Discord webhook, or leave that field blank to use the available account webhook. Choose Minimum score from 30 to 100 and Minimum drop % from 10 to 100, then Save settings. Both thresholds must be crossed for an alert.
Check & send now performs an immediate check when alerts are enabled. Matching alerts are sent once per creator per month. A message that all matching drops have already been alerted is therefore expected on a repeat run. Automatic checks require the retention check to be scheduled through the site’s Cron Manager; saving settings alone is not a schedule.
Troubleshooting
Monthly data is needed means imports are missing. A New / no baseline creator needs previous-month evidence. No matching queue rows may simply reflect restrictive filters or resolved actions.
Enable retention alerts before sending, and configure a webhook before enabling them. No creators cross both thresholds is a successful check with nothing to send. An outside-scope error requires an assignment/access review rather than a different search term.
Frequently asked questions
Does a high risk score mean the creator is leaving?
No. It highlights recorded performance decline and activity signals. Read the factors and discuss the context.
Why is a creator labelled New / no baseline?
There is no prior-month data available for the comparison.
Does Save plan notify the creator?
It saves the shared internal attention workflow. It is not a creator message or Discord alert action.
Why does Check & send now send nothing on a second attempt?
The same matching creator is alerted only once per month, or no creators cross both thresholds.
Can managers change alert thresholds?
No. They can maintain visible creators’ attention plans, while the owner manages alerts.
Still stuck? Note the page name, the step you tried and the exact message shown. Managers and sub-users should contact their network owner; owners should contact their platform administrator. Never include passwords, webhook URLs, recovery links or private identity documents in a general support message.