---
name: mm-analytics
description: >
  Analyze website SEO performance (Google Search Console) and ad campaign
  performance (Google Ads, Meta Ads) for a client. Produces insights and
  recommendations for content and marketing strategy.
  Use when: "analytics", "วิเคราะห์", "ดูสถิติ", "performance report",
  "seo report", "ad performance", "mm analytics", "ดูผล".
---

# mm-analytics

Analyzes SEO and ad performance for a client, produces insights and recommendations.

## When This Must Trigger

- "analytics", "วิเคราะห์", "ดูสถิติ", "performance report"
- "seo report", "ad performance", "mm analytics", "ดูผล"
- "วิเคราะห์เว็บ", "ดู traffic", "ดู campaign"

## Scripts Location

All scripts are in `~/Gitea/moreminimore-service-system/scripts/`:
- `mm_gsc.py` — Google Search Console analytics
- `mm_google_ads.py` — Google Ads analytics
- `mm_meta_ads.py` — Meta Ads analytics

## Process

### Step 1: Determine client and analysis type

```
ต้องการวิเคราะห์อะไร?

1. 🌐 SEO (Google Search Console) — traffic, queries, rankings
2. 📢 Google Ads — campaign performance, keywords
3. 📱 Meta Ads — Facebook/Instagram ad performance
4. 📊 ทั้งหมด — comprehensive report
```

### Step 2: Run analytics scripts

#### 2a. SEO Analytics (GSC)

```bash
cd ~/Gitea/moreminimore-service-system/scripts
python3 mm_gsc.py --client <client-id> --days 90
```

Output:
- Top queries by impressions, clicks, CTR
- Top pages by traffic
- Position distribution (top 3, 4-10, 11-20, 21+)
- Traffic trends (increasing/decreasing)
- Content gap opportunities

#### 2b. Google Ads Analytics

```bash
cd ~/Gitea/moreminimore-service-system/scripts
python3 mm_google_ads.py --client <client-id> campaigns
python3 mm_google_ads.py --client <client-id> keywords
python3 mm_google_ads.py --client <client-id> stats --days 30
```

Output:
- Active campaigns with performance metrics
- Keyword performance (impressions, clicks, CTR, CPC, QS)
- Campaign stats with cost per conversion

#### 2c. Meta Ads Analytics

```bash
cd ~/Gitea/moreminimore-service-system/scripts
python3 mm_meta_ads.py --client <client-id> campaigns
python3 mm_meta_ads.py --client <client-id> stats --days 30
```

Output:
- Active campaigns with objectives and budgets
- Performance stats (impressions, clicks, spend, reach, CTR, CPC)

### Step 3: Analyze and produce insights

After collecting data, analyze for:

**SEO Insights:**
- Queries with high impressions but low CTR → title/description optimization
- Queries ranking 4-10 → quick win opportunities
- Pages losing traffic → content refresh needed
- Content gaps → new article topics
- GEO opportunities → AI search optimization

**Ad Insights:**
- Campaigns with high CPA → optimize targeting/creative
- Keywords with low QS → improve ad relevance
- Ad fatigue signals → refresh creative
- Budget allocation → shift spend to better performers

**Cross-channel Insights:**
- SEO queries that match ad keywords → organic vs paid synergy
- Content topics that perform well in ads → create organic content
- Landing page performance → conversion optimization

### Step 4: Generate recommendations

Produce actionable recommendations:

```
📋 Analytics Report: [Client Name]
📅 Period: [date range]

━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
🌐 SEO Performance
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

Top Queries: [list]
Top Pages: [list]
Traffic Trend: [↑/↓/→]

⚡ Quick Wins:
1. [Query] — position [X], [Y] impressions → optimize title
2. [Page] — losing traffic → refresh content

📝 Content Opportunities:
1. [Topic] — high search volume, no content → write article
2. [Topic] — ranking 11-20 → expand existing content

━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
📢 Google Ads Performance
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

Campaigns: [N] active
Total Spend: ฿[X]
Avg CPA: ฿[X]
ROAS: [X]x

⚡ Recommendations:
1. [Campaign] — high CPA → adjust targeting
2. [Keyword] — low QS → improve ad relevance

━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
📱 Meta Ads Performance
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

Campaigns: [N] active
Total Spend: ฿[X]
Reach: [X]
CTR: [X]%

⚡ Recommendations:
1. [Campaign] — low CTR → refresh creative
2. [Ad Set] — high frequency → new audience

━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
📋 Next Actions
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

Content:
1. Write article about [topic] (from SEO gap)
2. Refresh [page] (losing traffic)

Ads:
1. Optimize [campaign] (high CPA)
2. Refresh [ad creative] (fatigue)

Strategy:
1. [Recommendation based on cross-channel data]
```

### Step 5: Save report

Save analytics report to:
```
~/vault/60_Articles/<client-id>/analytics/YYYY-MM-DD-report.md
```

## Integration with Content Pipeline

The analytics feed back into the content pipeline:

1. **SEO gaps** → become new article topics (mm-article-idea-extract)
2. **Top performing content** → inform future content strategy
3. **Ad performance** → inform CTA and messaging for ad copy
4. **Cross-channel insights** → optimize overall marketing strategy

## Error Handling

If a script fails:
- Check if credentials are configured in clients-config.json
- Check if the API is enabled (Google Ads API, Meta Marketing API)
- Check if the access token is valid

```
❌ [Script] failed: [error]

สาเหตุที่เป็นไปได้:
- Credentials ไม่ถูกต้อง → ตรวจสอบ clients-config.json
- API ไม่ได้เปิดใช้งาน → เปิดใช้ใน Google Cloud / Meta Developer
- Token หมดอายุ → ต่ออายุ token

ต้องการ:
1. ลองใหม่
2. ข้าม → วิเคราะห์เฉพาะส่วนที่ใช้ได้
```

## Notes

- All scripts read from `~/vault/99_System/clients-config.json`
- GSC uses service account key from `global.google.service_account_key`
- Google Ads uses developer_token + gcloud ADC for authentication
- Meta Ads uses page_token or system_user_token from config
- Reports are saved to the client's vault folder for reference