> ## Documentation Index
> Fetch the complete documentation index at: https://social-b97141fb-auto-generate-llmstxt.mintlify.site/llms.txt
> Use this file to discover all available pages before exploring further.

# Raw Data Export

> Download row-level CSV snapshots for deeper ad‑hoc, cohort, and anomaly analysis beyond on-screen dashboard visualizations.

<Info>
  Use Raw Data Export when you need to move from dashboard trends to precise, joinable event rows—enabling custom pivots, cohort retention, feature adoption deep-dives, and external BI modeling.
</Info>

## What You Can Answer

<Tabs>
  <Tab title="Engagement Spike Root Cause">Which users & post types drove yesterday’s reaction surge?</Tab>
  <Tab title="Long Tail Visibility">Which older posts resurfaced (impressions without recent creation)?</Tab>
  <Tab title="Format Performance">Did video posts sustain higher reaction/post ratios post-campaign?</Tab>
  <Tab title="Role Analysis">Are moderators or power users driving disproportionate commentary?</Tab>
  <Tab title="Cohort Retention">Do posts created in week N continue generating impressions weeks later?</Tab>
</Tabs>

## Export Workflow

<Steps>
  <Step title="Select Date Range">Adjust widget timeframe (e.g., last 7 / 30 days). Smaller ranges export faster & reduce row truncation risk.</Step>
  <Step title="Open Widget Menu">In the target widget header (Daily Engagements, Post Types, Reach & Impressions) click the Download icon.</Step>
  <Step title="Confirm Email">Pre-filled from your account; update if you need to CC an alias that’s allowed.</Step>
  <Step title="Generate Snapshot">Click Generate to queue export—toast confirmation appears.</Step>
  <Step title="Retrieve Email Links">Email contains signed URLs for each CSV; links valid 24h.</Step>
  <Step title="Archive & Version">Store files with a date-stamped naming convention (e.g., posts\_2025-08-17.csv).</Step>
  <Step title="Join & Analyze">Combine on shared keys (`postId`, user IDs, `targetId`) in your BI / spreadsheet tool.</Step>
</Steps>

## Available Files

| Widget              | File(s)         | Granularity                           | Primary Keys                       | Typical Question                         |
| ------------------- | --------------- | ------------------------------------- | ---------------------------------- | ---------------------------------------- |
| Daily Engagements   | posts.csv       | One row per post created on date      | postId, posterUserId               | Which creators spiked output?            |
| Daily Engagements   | comments.csv    | One row per comment/reply on date     | commentId, parentCommentId, postId | Where did discussion concentrate?        |
| Daily Engagements   | reactions.csv   | Each reaction event                   | reactionId, postId, reactorUserId  | Are reactions broad or concentrated?     |
| Post Types          | post\_types.csv | One row per post (lifecycle snapshot) | postId                             | Are video posts share increasing?        |
| Reach & Impressions | reach.csv       | Post-day tuple (unique viewers)       | postId + date                      | Which dormant posts regained visibility? |
| Reach & Impressions | impressions.csv | Post-day tuple (total renders)        | postId + date                      | Frequency vs reach divergence?           |

<Info>All files capped at 10,000 rows. Split large periods or narrow filters (community, post type) when approaching limit.</Info>

## Column Reference

### Common Core Fields

| Column                                         | Description                                             |
| ---------------------------------------------- | ------------------------------------------------------- |
| date                                           | UTC ISO 8601 day of the event/render                    |
| postId / commentId / reactionId                | Unique event identifiers                                |
| posterUserId / commenterUserId / reactorUserId | Originating user IDs                                    |
| targetType                                     | Location container (community, userFeed, etc.)          |
| targetId                                       | ID of target container                                  |
| postType                                       | Content classification (text, image, video, poll, etc.) |
| deleted                                        | Boolean flag (excludes rows if deleted before export)   |

### File Specific Highlights

<AccordionGroup>
  <Accordion title="posts.csv">Includes `postText`, `postType`, tag arrays, and deletion flags for content mining & classification.</Accordion>
  <Accordion title="comments.csv">Contains `parentCommentId` enabling thread reconstruction & depth analysis.</Accordion>
  <Accordion title="reactions.csv">Lightweight engagement events; join to posts for reaction intensity per post type.</Accordion>
  <Accordion title="post_types.csv">One row per post with full creator, timing, and type metadata for mix / composition trends.</Accordion>
  <Accordion title="reach.csv">Each row = post + day; reflects distinct viewer exposure irrespective of creation date.</Accordion>
  <Accordion title="impressions.csv">Frequency dimension per post + day; compare with reach to compute repetition (Impressions/Reach).</Accordion>
</AccordionGroup>

## Data Freshness & Snapshot Semantics

| Aspect        | Behavior                                                               |
| ------------- | ---------------------------------------------------------------------- |
| Point-in-Time | Export reflects current state at generation moment.                    |
| Edits         | Latest edited text retained (no historical versions).                  |
| Deletions     | Hard-deleted content omitted to respect policy.                        |
| Late Events   | Events arriving after generation won’t appear (re-export if critical). |
| Timezone      | All timestamps normalized to UTC (convert downstream as needed).       |

## Governance & Limits

| Area        | Policy                                                                          |
| ----------- | ------------------------------------------------------------------------------- |
| Row Cap     | 10,000 rows per CSV (enforced server-side).                                     |
| Frequency   | Manual—no scheduling yet (automation roadmap).                                  |
| Security    | Links are signed & expire in 24h; share internally only.                        |
| Permissions | Respects dashboard visibility scopes; if you can’t see it, you can’t export it. |
| Format      | UTF-8 CSV, comma-delimited, quoted where needed.                                |

<Info>For larger ongoing pipelines plan migration to forthcoming API / warehouse sync instead of stitching multiple manual exports.</Info>

## Joining & Derived Metrics

| Goal                    | Join Strategy                                       | Derived Formula                            |
| ----------------------- | --------------------------------------------------- | ------------------------------------------ |
| Reactions per Post Type | reactions.csv → posts.csv on postId                 | count(reactions)/count(posts) per postType |
| Comment Depth           | comments.csv self-join via parentCommentId          | avg(reply\_depth)                          |
| Visibility Decay Curve  | impressions.csv grouped by days since post creation | impression\_count over lag day index       |
| Frequency Factor        | impressions.csv join reach.csv (postId+date)        | impressions / reach                        |
| Content Creator Mix     | posts.csv aggregated by posterUserRole              | % share by role                            |

## Troubleshooting

| Symptom                   | Likely Cause                    | Investigation                            | Resolution                                        |
| ------------------------- | ------------------------------- | ---------------------------------------- | ------------------------------------------------- |
| Missing expected post     | Deleted before export           | Check moderation logs                    | Re-run earlier (if recoverable) or note exclusion |
| Row count hits 10k        | Range too broad                 | Count estimate via shorter sample        | Split date range / add filter                     |
| Stale email arrival       | Mail delay / processing backlog | Check status page & timestamp            | Wait; re-trigger if >30m                          |
| Empty CSV                 | No events in range              | Validate widget chart shows data         | Adjust date or remove restrictive filters         |
| Data mismatch vs chart    | Filter drift after export       | Confirm filters at generation vs current | Regenerate with correct filters                   |
| Character encoding issues | Non-UTF tool import settings    | Inspect raw file header                  | Force UTF-8 in BI tool                            |

## Best Practices

<AccordionGroup>
  <Accordion title="Narrow First">Export smallest viable slice; expand gradually to avoid hitting limits.</Accordion>
  <Accordion title="Version Files">Use ISO date naming; store in a controlled folder for reproducibility.</Accordion>
  <Accordion title="Pre-Join Profiles">Create small dimension tables (user roles, community metadata) and join post-import.</Accordion>
  <Accordion title="Automate Later">Prototype insights manually before investing in pipeline automation.</Accordion>
  <Accordion title="Retention Modeling">Use impressions.csv to build decay curves for evergreen content strategy.</Accordion>
</AccordionGroup>

## FAQ

**Can I exceed 10k rows?** Not currently—split range or await API export.

**Are private communities included?** Yes, if your role grants visibility.

**Scheduled exports?** On roadmap—use calendar reminder interim.

**Historical text versions?** Not stored—only latest edit at generation time.