WisePilot automates data collection, performance scoring, quality refinement, and strategic recommendations so your dashboards stay current without manual intervention. This page explains the daily and weekly rhythms, what data refreshes when, and where manual action is still needed.
Daily Automation
Every night, WisePilot runs a sequence of automated jobs that collect data, aggregate metrics, score assets, and surface insights. By the time you open the dashboard each morning, everything is up to date.
What Happens Overnight
| Order | What Runs | What It Produces |
|---|
| 1 | Google data import | Yesterday’s GA4 pageviews, users, and engagement + GSC impressions, clicks, CTR, and positions are pulled into daily snapshots |
| 2 | HighLevel conversation sync | New conversations, appointments, opportunities, and agent prompt versions are synced. Conversation outcomes are classified. Daily agent snapshots are built. |
| 3 | Performance rollup | All event data (GTM events, Google snapshots, HighLevel metrics) is aggregated into a single daily performance record per asset |
| 4 | Objective assignment | AI analyzes each asset’s performance and assigns or updates optimization objectives with reasoning |
| 5 | Revision impact evaluation | For assets with recent content updates, the system compares before/after performance to measure whether the revision helped |
| 6 | Entity extraction | New or updated content is analyzed to extract entities (topics, keywords, concepts) and embed them for search and recommendations |
| 7 | Quota auto-scheduling | For workspaces with active quotas, content ideas are automatically generated and scheduled to fill gaps |
| 8 | Conversation quality scoring | HighLevel AI conversations are evaluated against quality rubrics and scored for response quality, goal progress, and lead handling |
Background Quality Refinement
Blog content goes through a two-pass generation process:
- Pass 1 (Immediate) — When you generate a blog post, the initial draft is delivered immediately
- Pass 2 (Background) — Within minutes, a background refinement pass runs automatically to improve readability, brand voice consistency, claim accuracy, and CTA effectiveness
Pass 2 runs continuously throughout the day — you don’t need to wait or trigger it. If you open a blog post and see improvements appearing, that’s the background refinement at work.
Pass 2 respects your edits. If you’ve modified a section after generation, the refinement pass skips that section to preserve your changes.
Weekly Automation
On Mondays, additional analysis jobs run to surface strategic insights:
| What Runs | What It Produces |
|---|
| Idea discovery | AI identifies content gaps — topics your ICPs care about that you haven’t covered, refresh candidates for aging content, and emerging opportunities. New suggestions appear in your Ideas inbox. |
| Weekly digest | A performance summary is compiled for each workspace — top performers, declining assets, scoring trends, and action recommendations |
| Growth loop | An observe → analyze → recommend cycle that identifies growth opportunities and sends recommendations via email |
| Agent prompt revision | For HighLevel agents with significant conversation volume (50+ in the past week), the system analyzes patterns and suggests prompt improvements |
Data Freshness Reference
Use this table to understand when each data source was last refreshed and what delay to expect.
| Data Source | Refresh Frequency | Expected Delay | Where You See It |
|---|
| GA4 metrics | Daily | ~24 hours (yesterday’s data) | Asset performance, scoring, reports |
| GSC keywords | Daily | ~24 hours | Keyword rankings, visibility scores |
| GTM events (CTA views/clicks) | Real-time | Minutes | CTA performance, attribution |
| HighLevel form submissions | Real-time (webhook) | Minutes | Attribution, conversion tracking |
| HighLevel conversations | Daily sync | Up to 24 hours | Conversation dashboard, agent metrics |
| HighLevel appointments | Daily sync | Up to 24 hours | Booking rates, outcome classification |
| Optimization scores | Daily (after rollup) | ~24 hours | Scoring dashboard, asset list |
| Objective assignments | Daily (after scoring) | ~24 hours | Asset detail, optimization tab |
| Revision impact | Daily (after scoring) | ~24 hours | Revision Impact report |
| Entity extraction | Daily | ~24 hours | Search, recommendations, gap analysis |
| Idea discovery | Weekly (Mondays) | Up to 7 days | Ideas inbox |
| Quality refinement (Pass 2) | Continuous | Minutes | Blog post content |
| Weekly digest | Weekly (Mondays) | Up to 7 days | Email, dashboard |
Freshness Status Indicators
Throughout the platform, freshness is shown as a badge:
| Badge | Meaning | Typical Cause |
|---|
| Fresh (green) | Data updated within the last 24 hours | Everything is working normally |
| Stale (yellow) | Data is 24–48 hours old | May be a temporary delay — check again tomorrow |
| Very Stale (red) | Data is more than 48 hours old | Something is broken — check Data Quality |
What’s Automatic vs. What You Do
Fully Automated (No Action Needed)
- Google data imports (GA4 + GSC snapshots)
- HighLevel conversation, appointment, and opportunity syncing
- Performance rollup and score calculations
- Objective assignment and revision impact evaluation
- Entity extraction and embedding
- Background content quality refinement (Pass 2)
- Weekly idea discovery and growth recommendations
- Conversation quality scoring
- Daily and nurture sequence snapshots
Triggered By You, Then Automated
| You Do | WisePilot Automates |
|---|
| Generate a blog post | Background quality refinement runs within minutes |
| Set a content quota | Ideas are auto-generated and scheduled to fill it |
| Publish content to WordPress | Attribution tracking begins, performance data flows in |
| Connect Google accounts | Daily imports begin the next cycle |
| Connect HighLevel | Conversation sync begins the next cycle |
| Apply a content revision | Impact is measured against the pre-revision baseline |
Manual Actions We Recommend
These actions benefit from human judgment and can’t be fully automated:
| Action | Frequency | Why |
|---|
| Review idea suggestions | Weekly | Accept or dismiss AI-discovered topics based on your strategic priorities |
| Check Data Quality Console | Weekly | Catch attribution leaks or stale connections early — see Data Quality |
| Review conversation transcripts | As needed | AI scores conversations, but reviewing edge cases reveals prompt improvement opportunities |
| Update scoring weights | Quarterly | As your goals shift (awareness → conversion), adjust which metrics matter most |
| Refresh brand guidelines | When brand evolves | Updated guidelines immediately improve all future generation |
Backfill and Manual Sync
When you first connect an integration, there’s no historical data yet. WisePilot handles this differently per source:
| Source | Initial Behavior | Backfill Available? |
|---|
| GA4 | Next daily cycle imports yesterday’s data | Yes — you can backfill up to 90 days of historical data from the integration settings |
| GSC | Next daily cycle imports yesterday’s data | Yes — same 90-day backfill |
| HighLevel conversations | Next daily cycle syncs the last 24 hours | Yes — use the conversation import wizard to backfill historical conversations |
| GTM events | Events flow in real-time as they happen | No backfill — only events that fire after setup are captured |
| WordPress content | Manual import via bulk backfill | Yes — import existing posts/pages from the WordPress integration page |
After connecting Google accounts or HighLevel, run a backfill to populate historical data immediately rather than waiting for it to accumulate day by day. This gives you meaningful scores and reports from day one.
Troubleshooting Automation
Scores aren’t updating
- Check the Data Quality Console for freshness status
- If the daily rollup is stale, scores can’t update — this is the most common cause
- If rollup is fresh but scores seem wrong, check Scoring configuration
Idea suggestions aren’t appearing
- Discovery runs weekly on Mondays. If you just connected, wait until the following Monday.
- Discovery depends on entity extraction — if your content library is new, there may not be enough data to identify gaps yet.
- Check that you have at least one ICP configured — discovery uses ICPs to determine relevant topics.
HighLevel data seems outdated
- Conversations sync daily. If a conversation happened today, it will appear tomorrow morning.
- Real-time form submissions arrive immediately via webhook — check the Data Quality Console to verify the webhook is active.
- Appointments and opportunities sync daily alongside conversations.
Blog quality refinement didn’t run
- Pass 2 runs every few minutes. If you don’t see improvements within 15 minutes, the post may not have triggered refinement (it skips posts that already pass quality thresholds).
- If you edited the post before refinement ran, those sections are intentionally skipped to preserve your changes.
