Analytics in GenieHelper are layered by purpose: per-post engagement is separate from period revenue rollups, which are separate from account-level growth curves. This separation keeps the data model clean and makes queries fast — you are never aggregating per-post data to get a revenue summary.
Analytics data is sourced from platform scrapes via Stagehand, not from platform APIs. This means data freshness is tied to scrape frequency (every 6 hours by default for the scrape_scheduler) and to HITL scrape sessions when platforms block headless Chrome.
Earnings history
Earnings are tracked in creator_earnings as period-by-period revenue records. Each record represents one revenue period for one platform and one content type.
What earnings records contain
| Field | Description |
|---|
| Period | Date range (daily, weekly, or monthly depending on platform reporting) |
| Platform | Which platform this revenue came from (OnlyFans, Fansly, etc.) |
| Content type | Subscription revenue, PPV, tips, custom content, referrals |
| Gross amount | Platform-reported gross before fees |
| Net amount | After platform percentage and any applicable taxes |
| Currency | ISO currency code |
Revenue breakdown
Creators can slice earnings by:
- Platform — which platforms are generating the most revenue
- Content type — whether subscriptions, PPV, or tips are the primary driver
- Time period — daily, weekly, monthly, or custom date range
- Campaign attribution — which campaigns drove measurable revenue lift (via
campaign_platform_performance)
Creator goals
Creator goals in creator_goals let creators set revenue, subscriber, or engagement targets with automatic progress tracking.
Goal types
| Type | Examples |
|---|
| Revenue | Hit 5,000/monthnet,reach50,000 annual gross |
| Subscriber count | Reach 1,000 active OnlyFans subscribers, maintain Fansly subscriber count above 500 |
| Engagement | Achieve 40% average open rate on broadcasts, hit 20 PPV purchases per week |
Progress tracking
Goal progress is calculated automatically by comparing the goal target against the most recent relevant data:
- Revenue goals compare against
creator_earnings sum for the goal period
- Subscriber count goals compare against the latest
platform_stats_snapshots subscriber count
- Engagement goals compare against
post_performance_snapshots averages
Progress is surfaced in the AnalyticsAdvancedPanel (right wing) and can be called up in chat. The agent monitors goal progress and can proactively alert the creator when a goal is at risk or has been reached.
The agent can create and update creator goals via conversation. Say “set a goal to reach 1,000 OnlyFans subscribers by June” and the agent will create the record and start tracking it.
platform_stats_snapshots — one record per platform per scrape cycle.
What each snapshot contains
| Field | Description |
|---|
| Platform | Which platform this snapshot belongs to |
| Follower count | Total followers at snapshot time |
| Subscriber count | Total active paid subscribers |
| Churn rate | Percentage of subscribers who cancelled in the snapshot period |
| New subscribers | Net new subscribers since the last snapshot |
| Snapshot timestamp | When this data was collected |
AnalyticsAdvancedPanel renders this as a time series chart. The agent can identify growth inflection points and correlate them with campaign activity or posting frequency changes.
platform_stats_snapshots is append-only. The current state of a platform connection lives on the platform_connections record; historical trends live in snapshots. Never read current subscriber counts from snapshots — always use the connection record.
Campaign ROI
Campaign performance is broken down per platform in campaign_platform_performance.
Per-campaign metrics
| Metric | Source |
|---|
| Impressions | Scraped from platform post analytics |
| Clicks | Link click counts from platform reporting |
| Conversions | New subscribers or PPV purchases attributed to the campaign |
| Revenue | Gross revenue attributed within the campaign attribution window |
| ROI | (Revenue − Cost) / Cost, where cost is entered manually by the creator |
AnalyticsAdvancedPanel and can be queried in chat. The agent uses campaign performance data to recommend which content types and platforms to prioritize for future campaigns.
Post performance snapshots
Per-post engagement is tracked in post_performance_snapshots. Each record captures the live metrics for a specific post at a specific point in time.
| Field | Description |
|---|
| Post reference | Link to the content_posts record |
| Platform | Which platform this performance data is from |
| Snapshot timestamp | When this data was collected |
| Views | Total view count at snapshot time |
| Likes | Total like count |
| Comments | Total comment count |
| Shares | Total share count (where platform exposes it) |
| Link clicks | CTA link click count (where trackable) |
| Revenue | PPV revenue attributed to this post |
How snapshots are collected
Post performance data is collected by the scrape_post_performance handler in operations/publish.js. It uses Stagehand to navigate to each post’s analytics view on the platform and extract the current metrics.
After writing the snapshot record, the handler triggers the “PUB: Snapshot Evaluated” Directus Flow, which can fire downstream automations — for example, promoting a high-performing post to a segment broadcast.
The scrape_post_performance handler was added on 2026-03-15 and is registered on the scrape-jobs BullMQ queue, not media-jobs. This keeps browser automation jobs separate from FFmpeg processing jobs.
Key collections reference
| Collection | Layer | Purpose |
|---|
creator_earnings | Period rollup | Revenue by platform and content type per period |
creator_goals | Targets | Revenue, subscriber, and engagement goals with progress tracking |
platform_stats_snapshots | Account time series | Historical follower count, subscriber count, churn rate |
campaign_platform_performance | Campaign attribution | Per-platform revenue, clicks, conversions per campaign |
post_performance_snapshots | Post-level engagement | Live metrics per post per platform at snapshot time |
Analytics data flow