Analytics Reporter - Exemple complet

Créez un agent IA qui se connecte à Google Analytics, analyse le trafic de votre site web, génère des insights et envoie des rapports quotidiens automatisés à votre équipe.

Ce que vous allez construire

Un agent de reporting analytics qui :

Se connecte à Google Analytics 4 (GA4)
Interroge les métriques de trafic et le comportement des utilisateurs
Identifie les tendances et les anomalies
Génère des rapports quotidiens/hebdomadaires
Crée des recommandations actionnables
Envoie les rapports par Slack ou email

Temps nécessaire : 45-60 minutes

Prérequis

Avant de commencer, vous avez besoin de :

Un compte TeamDay avec une organisation
Une propriété Google Analytics 4 avec des données
Un projet Google Cloud (pour OAuth)
Une clé API Anthropic
(Optionnel) Un workspace Slack pour les notifications

Vue d’ensemble de l’architecture

graph LR
    A[Scheduled Mission] --> B[Analytics Agent]
    B --> C[Google Analytics MCP]
    C --> D[GA4 API]
    B --> E[Data Analysis]
    E --> F[Report Generation]
    F --> G[Slack/Email]
    B --> H[Insights Agent]
    H --> I[Recommendations]

Étape 1 : Configuration Google Cloud

1.1 Créer un projet Google Cloud

Rendez-vous sur Google Cloud Console
Créez un nouveau projet : “TeamDay Analytics”
Notez votre Project ID : teamday-analytics-12345

1.2 Activer l’API Google Analytics

# Avec gcloud CLI
gcloud services enable analyticsdata.googleapis.com
gcloud services enable analytics.googleapis.com

# Ou via Cloud Console :
# APIs & Services → Enable APIs → Rechercher "Google Analytics Data API"

1.3 Créer des identifiants OAuth 2.0

Créer un client OAuth :

APIs & Services → Credentials
Create Credentials → OAuth client ID
Type d’application : Web application
Nom : “TeamDay Analytics Reporter”
URIs de redirection autorisés :
- https://app.teamday.app/auth/google/callback
- http://localhost:3000/auth/google/callback (pour les tests)
Cliquez sur Create
Téléchargez les identifiants JSON

Sauvegardez ces valeurs :

{
  "client_id": "123456789-abc.apps.googleusercontent.com",
  "client_secret": "GOCSPX-abcdef123456",
  "redirect_uris": ["https://app.teamday.app/auth/google/callback"]
}

1.4 Obtenir un Refresh Token

Vous avez besoin d’un refresh token pour accéder à GA sans autorisation manuelle :

# Installer la bibliothèque Google OAuth
npm install -g google-auth-library

# Créer un script d'authentification
cat > get-refresh-token.js << 'EOF'
const { OAuth2Client } = require('google-auth-library');
const readline = require('readline');

const client = new OAuth2Client(
  'YOUR_CLIENT_ID',
  'YOUR_CLIENT_SECRET',
  'http://localhost:3000/auth/google/callback'
);

const SCOPES = [
  'https://www.googleapis.com/auth/analytics.readonly',
  'https://www.googleapis.com/auth/analytics'
];

const authUrl = client.generateAuthUrl({
  access_type: 'offline',
  scope: SCOPES,
});

console.log('Authorize this app by visiting:\n', authUrl);

const rl = readline.createInterface({
  input: process.stdin,
  output: process.stdout,
});

rl.question('Enter the code from that page here: ', async (code) => {
  rl.close();
  try {
    const { tokens } = await client.getToken(code);
    console.log('\nRefresh Token:', tokens.refresh_token);
    console.log('\nSave this token securely!');
  } catch (error) {
    console.error('Error getting token:', error);
  }
});
EOF

# Exécuter le script
node get-refresh-token.js

Suivez les instructions :

Visitez l’URL d’autorisation
Accordez les permissions
Copiez le code de l’URL de redirection
Collez-le dans le terminal
Sauvegardez le refresh token : 1//abc123def456...

1.5 Obtenir l’ID de propriété GA4

Trouvez votre ID de propriété GA4 :

Rendez-vous sur Google Analytics
Admin → Property Settings
Copiez l’ID de propriété : 123456789

Ou trouvez-le dans l’URL :

https://analytics.google.com/analytics/web/#/p123456789/...
                                             ^^^^^^^^^^

Étape 2 : Configuration de l’organisation TeamDay

2.1 Ajouter les clés API

Stockez vos identifiants dans TeamDay :

teamday auth login
teamday secrets set ANTHROPIC_API_KEY "sk-ant-api03-..."
teamday secrets set GA_CLIENT_ID "123456789-abc.apps.googleusercontent.com"
teamday secrets set GA_CLIENT_SECRET "GOCSPX-abcdef123456"
teamday secrets set GA_REFRESH_TOKEN "1//abc123def456..."
teamday secrets set GA_PROPERTY_ID "123456789"

Ou via l’API :

curl -X POST https://us.teamday.ai/api/secrets/set \
  -H "Authorization: Bearer $TEAMDAY_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{
    "organizationId": "org_xxx",
    "secrets": {
      "GA_CLIENT_ID": "123456789-abc.apps.googleusercontent.com",
      "GA_CLIENT_SECRET": "GOCSPX-abcdef123456",
      "GA_REFRESH_TOKEN": "1//abc123def456...",
      "GA_PROPERTY_ID": "123456789"
    }
  }'

2.2 Créer un Space

Créez un space pour le travail d’analytics :

teamday spaces create "Analytics Reports"

# Obtenir l'ID du space
teamday spaces list
# Output: cs_analytics123...

Étape 3 : Installer Google Analytics MCP

3.1 Configuration du serveur MCP

Le serveur MCP Google Analytics permet à votre agent d’interroger les données GA4 :

Configuration :

{
  "mcpServers": {
    "google-analytics": {
      "command": "npx",
      "args": ["-y", "@modelcontextprotocol/server-google-analytics"],
      "env": {
        "GA_CLIENT_ID": "${GA_CLIENT_ID}",
        "GA_CLIENT_SECRET": "${GA_CLIENT_SECRET}",
        "GA_REFRESH_TOKEN": "${GA_REFRESH_TOKEN}",
        "GA_PROPERTY_ID": "${GA_PROPERTY_ID}"
      }
    }
  }
}

3.2 Installation via l’interface

Ouvrez votre space
Settings → MCP Plugins
Cliquez sur “Install Plugin”
Recherchez “Google Analytics”
Configuration auto-remplie depuis les secrets
Cliquez sur “Install & Test”

3.3 Installation via l’API

curl -X POST https://us.teamday.ai/api/plugins/install \
  -H "Authorization: Bearer $TEAMDAY_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{
    "spaceId": "cs_analytics123",
    "plugin": "google-analytics",
    "config": {
      "mcpServers": {
        "google-analytics": {
          "command": "npx",
          "args": ["-y", "@modelcontextprotocol/server-google-analytics"],
          "env": {
            "GA_CLIENT_ID": "${GA_CLIENT_ID}",
            "GA_CLIENT_SECRET": "${GA_CLIENT_SECRET}",
            "GA_REFRESH_TOKEN": "${GA_REFRESH_TOKEN}",
            "GA_PROPERTY_ID": "${GA_PROPERTY_ID}"
          }
        }
      }
    }
  }'

3.4 Vérifier l’installation MCP

Testez la connexion :

Vous : Test the Google Analytics connection

Agent : Testing Google Analytics MCP...

✅ Authentication successful
✅ Connected to property: 123456789
✅ Available metrics: 45
✅ Date range: 2023-01-01 to 2024-12-09

Ready to query analytics data!

Étape 4 : Créer l’agent Analytics

4.1 Créer l’agent principal

curl -X POST https://us.teamday.ai/api/v1/agents \
  -H "Authorization: Bearer $TEAMDAY_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{
    "name": "Analytics Reporter",
    "model": "claude-sonnet-4-6",
    "visibility": "organization",
    "organizationId": "org_xxx"
  }'

Sauvegardez l’ID de l’agent : ag_analytics456

4.2 Configuration du System Prompt

# Analytics Reporter System Prompt

You are an expert data analyst specializing in web analytics and user behavior.

## Your Responsibilities

1. **Data Collection**
   - Query Google Analytics for relevant metrics
   - Pull data for specified time periods
   - Handle date ranges and comparisons

2. **Analysis**
   - Identify trends and patterns
   - Spot anomalies and significant changes
   - Calculate growth rates and percentages
   - Compare periods (week-over-week, month-over-month)

3. **Insights Generation**
   - Explain what the data means
   - Identify opportunities and concerns
   - Provide context for changes
   - Highlight actionable insights

4. **Report Creation**
   - Structure reports clearly
   - Use appropriate visualizations (describe them)
   - Prioritize most important findings
   - Include recommendations

## Key Metrics to Track

**Traffic Metrics:**
- Total users and sessions
- New vs returning users
- Pageviews and unique pageviews
- Bounce rate
- Average session duration

**Acquisition:**
- Traffic sources (organic, direct, referral, social)
- Top referring sites
- Campaign performance
- Search keywords (if available)

**Behavior:**
- Top pages by views
- Top landing pages
- Exit pages
- User flow patterns

**Conversions:**
- Goal completions
- Conversion rates
- E-commerce metrics (if applicable)
- Form submissions

**Technical:**
- Device breakdown (desktop, mobile, tablet)
- Browser and OS distribution
- Page load times
- Error rates

## Analysis Framework

For each metric:
1. Current value
2. Comparison to previous period
3. Trend direction (↑↓→)
4. Significance of change
5. Possible explanations

## Report Format

Structure reports as:

### 📊 Analytics Report - [Date Range]

**Summary**
High-level overview (2-3 sentences)

**Key Findings**
- Most important insights
- Significant changes
- Opportunities or concerns

**Traffic Overview**
- User metrics
- Session metrics
- Engagement metrics

**Acquisition Analysis**
- Traffic sources breakdown
- Top performers
- Changes from last period

**Content Performance**
- Top pages
- Engagement by content type
- User journey insights

**Recommendations**
- Actionable next steps
- Areas to investigate
- Optimization opportunities

Use emojis for visual clarity: 📈 (growth), 📉 (decline), ⚠️ (concern), ✨ (opportunity)

Be data-driven but explain insights in plain language.

4.3 Définir les permissions de l’agent

curl -X PATCH https://us.teamday.ai/api/v1/agents/ag_analytics456 \
  -H "Authorization: Bearer $TEAMDAY_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{
    "permissions": {
      "readFiles": true,
      "writeFiles": true,
      "executeCode": false,
      "commitToGit": false,
      "accessSecrets": false
    }
  }'

Étape 5 : Créer un sous-agent Insights (optionnel)

5.1 Créer un sous-agent spécialisé

Pour une analyse plus approfondie, créez un agent d’insights :

curl -X POST https://us.teamday.ai/api/v1/agents \
  -H "Authorization: Bearer $TEAMDAY_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{
    "name": "Insights Analyzer",
    "systemPrompt": "You analyze analytics data for trends and insights...",
    "model": "claude-sonnet-4-6",
    "visibility": "organization",
    "organizationId": "org_xxx",
    "type": "subagent"
  }'

5.2 Lier à l’agent principal

curl -X PATCH https://us.teamday.ai/api/v1/agents/ag_analytics456 \
  -H "Authorization: Bearer $TEAMDAY_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{
    "subAgents": ["ag_insights789"]
  }'

Étape 6 : Exemples d’utilisation

6.1 Requête analytics de base

Vous : What were our top pages yesterday?

Agent : Querying Google Analytics for yesterday's data...

## Top Pages - December 8, 2024

1. **/blog/getting-started**
   - 1,247 views
   - Avg time: 3m 24s
   - Bounce rate: 42%

2. **/docs/api-reference**
   - 892 views
   - Avg time: 5m 12s
   - Bounce rate: 28%

3. **/pricing**
   - 634 views
   - Avg time: 2m 08s
   - Bounce rate: 55%

4. **/features**
   - 521 views
   - Avg time: 1m 45s
   - Bounce rate: 62%

5. **/** (Homepage)
   - 489 views
   - Avg time: 1m 22s
   - Bounce rate: 48%

**Insights:**
- Getting started guide remains top content
- API docs show high engagement (5+ min sessions)
- Pricing page has high bounce rate - consider optimization

6.2 Comparaison de trafic

Vous : Compare this week's traffic to last week

Agent : Analyzing week-over-week traffic data...

## Weekly Traffic Comparison

**This Week (Dec 2-8)** vs **Last Week (Nov 25-Dec 1)**

### Overall Traffic
- **Total Users:** 12,453 ↑ (+18.2%)
  - Previous: 10,536
  - Change: +1,917 users

- **Sessions:** 15,892 ↑ (+16.7%)
  - Previous: 13,618
  - Change: +2,274 sessions

- **Pageviews:** 42,156 ↑ (+22.3%)
  - Previous: 34,478
  - Change: +7,678 views

### Engagement Metrics
- **Avg Session Duration:** 3m 42s ↑ (+8.3%)
  - Previous: 3m 25s
  - Improvement in engagement

- **Bounce Rate:** 45.2% ↓ (-3.1%)
  - Previous: 48.3%
  - Good improvement

- **Pages/Session:** 2.65 ↑ (+11.8%)
  - Previous: 2.37
  - Users exploring more

### Traffic Sources
1. **Organic Search:** 6,234 users ↑ (+28.4%)
2. **Direct:** 3,892 users ↑ (+12.1%)
3. **Referral:** 1,456 users ↑ (+5.3%)
4. **Social:** 871 users ↑ (+45.2%) 📈

**Key Findings:**
✨ Strong growth across all metrics
✨ Social traffic spike - recent campaign working well
✨ Improved engagement indicates better content resonance
⚠️ Monitor capacity - ensure site performance scales

6.3 Générer un rapport quotidien complet

Vous : Generate today's daily analytics report

Agent : Creating comprehensive analytics report for December 9, 2024...

---

# 📊 Daily Analytics Report
**December 9, 2024**

## Executive Summary

Strong performance today with 1,834 users (+12% vs yesterday). Organic traffic continues to grow, and engagement metrics are healthy. Notable spike in mobile traffic suggests mobile optimization efforts are paying off.

## 🎯 Key Highlights

✨ **New users increased 22%** - acquisition efforts working
📈 **Mobile traffic up 34%** - mobile-first strategy succeeding
⚠️ **Bounce rate on /pricing increased to 61%** - needs attention
✨ **Blog engagement up 18%** - content strategy effective

## 📈 Traffic Overview

### Total Traffic
- **Users:** 1,834 (↑12% vs yesterday)
- **New Users:** 1,245 (↑22% vs yesterday)
- **Sessions:** 2,367 (↑15% vs yesterday)
- **Pageviews:** 6,892 (↑18% vs yesterday)

### Engagement
- **Avg Session Duration:** 3m 48s (↑5%)
- **Pages/Session:** 2.91 (↑8%)
- **Bounce Rate:** 43.2% (↓2.1%)

## 🚀 Acquisition Breakdown

| Source | Users | % of Total | Change |
|--------|-------|------------|--------|
| Organic Search | 892 | 48.6% | ↑15% |
| Direct | 567 | 30.9% | ↑8% |
| Referral | 234 | 12.8% | ↑23% |
| Social | 98 | 5.3% | ↑45% |
| Email | 43 | 2.3% | ↑12% |

**Top Referrers:**
1. news.ycombinator.com - 89 users
2. reddit.com/r/webdev - 56 users
3. dev.to - 34 users

## 📱 Device & Technology

### Device Category
- **Desktop:** 1,102 users (60.1%)
- **Mobile:** 652 users (35.5%) ↑34%
- **Tablet:** 80 users (4.4%)

### Browser
- Chrome: 62.3%
- Safari: 24.1%
- Firefox: 8.9%
- Edge: 3.2%
- Other: 1.5%

## 📄 Content Performance

### Top Pages
1. **/blog/api-best-practices** - 423 views
   - 4m 32s avg time
   - 32% bounce rate
   - 📈 Trending up

2. **/docs/getting-started** - 389 views
   - 5m 12s avg time
   - 28% bounce rate
   - → Steady

3. **/pricing** - 267 views
   - 1m 45s avg time
   - 61% bounce rate
   - ⚠️ High bounce rate

4. **/features/automation** - 198 views
   - 3m 21s avg time
   - 38% bounce rate

5. **/** (Homepage) - 187 views
   - 1m 33s avg time
   - 45% bounce rate

### Top Landing Pages
1. /blog/api-best-practices (327 sessions)
2. / (198 sessions)
3. /docs/getting-started (156 sessions)

## 🎯 User Behavior Insights

### User Flow Analysis
**Most Common Path:**
Homepage → Features → Pricing → Sign Up
- Conversion rate: 8.2%

**High Exit Pages:**
1. /pricing (23% exit rate)
2. /contact (18% exit rate)

### Session Depth
- 1 page: 43.2%
- 2-3 pages: 34.1%
- 4-6 pages: 15.8%
- 7+ pages: 6.9%

## 💡 Recommendations

### High Priority
1. **Optimize /pricing page**
   - Current bounce rate: 61%
   - Add social proof or comparison table
   - Consider A/B test on CTA placement

2. **Capitalize on HackerNews traffic**
   - 89 referrals today
   - Create follow-up content
   - Engage in comments

### Medium Priority
1. **Mobile optimization working**
   - 34% increase in mobile traffic
   - Continue mobile-first approach
   - Test mobile checkout flow

2. **Blog performing well**
   - "API Best Practices" trending
   - Create more technical content
   - Add related article suggestions

### Monitoring
1. Watch organic growth trajectory
2. Track pricing page changes
3. Monitor mobile conversion rates

## 📊 Historical Context

**7-Day Trend:**
- Daily avg users: 1,654
- Today vs 7-day avg: ↑10.9%
- Consistent upward trend

**30-Day Comparison:**
- This week avg: 1,723 users/day
- Last 30 days avg: 1,498 users/day
- Growth: ↑15%

---

*Report generated by TeamDay Analytics Agent*
*Next report: December 10, 2024 at 9:00 AM UTC*

6.4 Demande d’analyse personnalisée

Vous : Analyze which traffic sources convert best

Agent : Analyzing conversion rates by traffic source...

## Conversion Analysis by Traffic Source

**Date Range:** Last 30 days
**Goal:** Sign-up conversion

### Conversion Rates by Source

| Source | Users | Conversions | Rate | AOV |
|--------|-------|-------------|------|-----|
| **Organic Search** | 15,234 | 892 | **5.9%** | $127 |
| **Email** | 1,245 | 156 | **12.5%** 📈 | $189 |
| **Referral** | 3,456 | 198 | **5.7%** | $145 |
| **Direct** | 8,923 | 445 | **5.0%** | $156 |
| **Social** | 2,134 | 67 | **3.1%** | $98 |
| **Paid Search** | 1,890 | 132 | **7.0%** | $134 |

### Key Insights

✨ **Email is your best converter** (12.5% rate)
- Highest rate and AOV
- Users are already engaged
- Recommendation: Increase email frequency, add nurture campaigns

📈 **Paid search over-performing organic** (7.0% vs 5.9%)
- Good keyword targeting
- Recommendation: Increase paid budget strategically

⚠️ **Social traffic underperforming** (3.1% rate)
- Low intent traffic
- Lowest AOV
- Recommendation: Better targeting, or focus on awareness vs conversion

### Revenue Impact

**Total Revenue (30 days):** $245,678

**By Source:**
1. Organic Search: $113,284 (46%)
2. Direct: $69,420 (28%)
3. Referral: $28,710 (12%)
4. Paid Search: $17,688 (7%)
5. Email: $29,484 (12%)
6. Social: $6,566 (3%)

**Recommendation:**
Despite lower conversion rate, organic drives most revenue due to volume. Focus on:
1. Maintaining organic rankings
2. Scaling email (best ROI)
3. Optimizing paid search (good performance)
4. Rethinking social strategy (low ROI)

Étape 7 : Automatisation et planification

7.1 Mission de rapport quotidien

Configurez des rapports quotidiens automatiques :

curl -X POST https://us.teamday.ai/api/missions \
  -H "Authorization: Bearer $TEAMDAY_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{
    "name": "Daily Analytics Report",
    "schedule": "0 9 * * *",
    "agentId": "ag_analytics456",
    "action": {
      "type": "execute",
      "prompt": "Generate today'\''s analytics report and post to #analytics Slack channel"
    },
    "organizationId": "org_xxx",
    "spaceId": "cs_analytics123"
  }'

7.2 Mission de résumé hebdomadaire

curl -X POST https://us.teamday.ai/api/missions \
  -H "Authorization: Bearer $TEAMDAY_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{
    "name": "Weekly Analytics Summary",
    "schedule": "0 9 * * 1",
    "agentId": "ag_analytics456",
    "action": {
      "type": "execute",
      "prompt": "Generate comprehensive weekly analytics report comparing to previous week, and email to [email protected]"
    },
    "organizationId": "org_xxx",
    "spaceId": "cs_analytics123"
  }'

Exemples de planification :

0 9 * * * - Tous les jours à 9h
0 9 * * 1 - Tous les lundis à 9h
0 */4 * * * - Toutes les 4 heures
0 9 1 * * - Le premier jour du mois à 9h

7.3 Mission de détection d’anomalies

Détectez les patterns de trafic inhabituels :

curl -X POST https://us.teamday.ai/api/missions \
  -H "Authorization: Bearer $TEAMDAY_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{
    "name": "Anomaly Detection",
    "schedule": "0 */6 * * *",
    "agentId": "ag_analytics456",
    "action": {
      "type": "execute",
      "prompt": "Check last 6 hours of traffic. Alert if traffic dropped >20% or spiked >50% compared to same time yesterday"
    },
    "organizationId": "org_xxx",
    "spaceId": "cs_analytics123"
  }'

Étape 8 : Intégration avec Slack (optionnel)

8.1 Installer Slack MCP

{
  "mcpServers": {
    "slack": {
      "command": "npx",
      "args": ["-y", "@modelcontextprotocol/server-slack"],
      "env": {
        "SLACK_BOT_TOKEN": "${SLACK_BOT_TOKEN}",
        "SLACK_CHANNEL": "#analytics"
      }
    }
  }
}

8.2 Envoyer un rapport à Slack

Vous : Send today's report to #analytics

Agent : Posting analytics report to Slack...

✅ Posted to #analytics
✅ Added chart visualizations
✅ Pinned message for visibility

View in Slack: https://yourworkspace.slack.com/...

Résultats attendus

Exemple de message Slack

📊 Daily Analytics Report - December 9, 2024

*Traffic Summary*
• Users: 1,834 (↑12%)
• Sessions: 2,367 (↑15%)
• Pageviews: 6,892 (↑18%)

*Key Highlights*
✨ New users up 22% - acquisition working
📈 Mobile traffic up 34% - optimization paying off
⚠️ Pricing bounce rate 61% - needs attention

*Top Pages Today*
1. /blog/api-best-practices (423 views)
2. /docs/getting-started (389 views)
3. /pricing (267 views)

*Recommendations*
• Optimize /pricing page (high bounce)
• Capitalize on HN traffic (89 referrals)
• Continue mobile-first approach

<View Full Report> | <Ask Follow-up>

Exemple de rapport par email

Subject: Daily Analytics Report - December 9, 2024

[Email HTML formaté avec graphiques et métriques]

Métriques clés, visualisations et insights
Lien vers le rapport complet dans TeamDay

Dépannage

Échec de l’authentification Google Analytics

Erreur :

Error: GA authentication failed (401 Unauthorized)

Solution :

Vérifiez que le refresh token est valide
Vérifiez que les identifiants OAuth sont corrects
Assurez-vous que l’API GA4 est activée dans GCP
Régénérez le refresh token s’il a expiré
Vérifiez que l’ID de propriété est correct

Aucune donnée retournée

Erreur :

Error: No data available for date range

Solution :

Vérifiez que la propriété contient des données pour les dates demandées
Vérifiez que l’ID de propriété est correct
Assurez-vous que l’utilisateur a accès en lecture à Analytics
Essayez une plage de dates plus courte
Vérifiez que la propriété est GA4 (pas Universal Analytics)

Le plugin MCP ne répond pas

Erreur :

Error: Google Analytics MCP timeout

Solution :

Vérifiez la connectivité réseau
Vérifiez que les APIs GCP sont activées
Vérifiez les limites de taux (quota)
Redémarrez le serveur MCP
Consultez les logs du plugin

Écarts dans les données

Problème : Les chiffres ne correspondent pas à l’interface GA

Solution :

Vérifiez que les plages de dates correspondent exactement
Vérifiez les paramètres de fuseau horaire
Tenez compte du délai de traitement des données (24-48h pour certaines métriques)
Utilisez l’API realtime pour les données actuelles
Vérifiez les filtres et segments

Configuration avancée

Configuration des métriques personnalisées

Définissez des métriques personnalisées à suivre :

{
  "customMetrics": {
    "signups": {
      "eventName": "sign_up",
      "type": "event"
    },
    "purchases": {
      "eventName": "purchase",
      "type": "event",
      "parameters": ["value", "currency"]
    },
    "engagement": {
      "eventName": "page_view",
      "conditions": {
        "session_duration": ">180"
      }
    }
  }
}

Modèles de rapports

Créez des modèles de rapports réutilisables :

{
  "templates": {
    "executive_summary": {
      "metrics": ["users", "sessions", "conversions"],
      "dimensions": ["source", "medium"],
      "dateRange": "last7days",
      "comparison": "previousPeriod"
    },
    "content_performance": {
      "metrics": ["pageviews", "avgSessionDuration", "bounceRate"],
      "dimensions": ["pagePath", "pageTitle"],
      "filters": {
        "pagePath": "startsWith:/blog"
      },
      "orderBy": "pageviews desc",
      "limit": 10
    }
  }
}

Seuils d’alerte

Configurez des alertes automatisées :

{
  "alerts": {
    "traffic_drop": {
      "metric": "users",
      "threshold": -20,
      "period": "1hour",
      "comparison": "same_time_yesterday",
      "action": "slack_notification"
    },
    "conversion_spike": {
      "metric": "conversions",
      "threshold": 50,
      "period": "1day",
      "action": "email_team"
    }
  }
}

Estimation des coûts

API Google Analytics :

Niveau gratuit : 25 000 requêtes/jour
Rapport quotidien typique : ~10 requêtes
Coût mensuel : $0 (dans le niveau gratuit)

Agent TeamDay :

Rapport quotidien : ~8 000 tokens en entrée, ~3 000 en sortie
Coût par rapport : ~$0.12
Mensuel (30 jours) : ~$3.60

Coût total estimé : ~$4/mois

Prochaines étapes

Maintenant que vous avez un agent de reporting analytics, explorez :

Exemple BigQuery Insights - Analysez les données historiques à grande échelle
Code Review Bot - Corrélez les changements de code avec le trafic
Serveurs MCP personnalisés - Créez des outils d’analytics personnalisés
Guide d’automatisation - Planification et déclencheurs avancés

Analytics Reporter - Exemple complet

Ce que vous allez construire

Prérequis

Vue d’ensemble de l’architecture

Étape 1 : Configuration Google Cloud

1.1 Créer un projet Google Cloud

1.2 Activer l’API Google Analytics

1.3 Créer des identifiants OAuth 2.0

1.4 Obtenir un Refresh Token

1.5 Obtenir l’ID de propriété GA4

Étape 2 : Configuration de l’organisation TeamDay

2.1 Ajouter les clés API

2.2 Créer un Space

Étape 3 : Installer Google Analytics MCP

3.1 Configuration du serveur MCP

3.2 Installation via l’interface

3.3 Installation via l’API

3.4 Vérifier l’installation MCP

Étape 4 : Créer l’agent Analytics

4.1 Créer l’agent principal

4.2 Configuration du System Prompt

4.3 Définir les permissions de l’agent

Étape 5 : Créer un sous-agent Insights (optionnel)

5.1 Créer un sous-agent spécialisé

5.2 Lier à l’agent principal

Étape 6 : Exemples d’utilisation

6.1 Requête analytics de base

6.2 Comparaison de trafic

6.3 Générer un rapport quotidien complet

6.4 Demande d’analyse personnalisée

Étape 7 : Automatisation et planification

7.1 Mission de rapport quotidien

7.2 Mission de résumé hebdomadaire

7.3 Mission de détection d’anomalies

Étape 8 : Intégration avec Slack (optionnel)

8.1 Installer Slack MCP

8.2 Envoyer un rapport à Slack

Résultats attendus

Exemple de message Slack

Exemple de rapport par email

Dépannage

Échec de l’authentification Google Analytics

Aucune donnée retournée

Le plugin MCP ne répond pas

Écarts dans les données

Configuration avancée

Configuration des métriques personnalisées

Modèles de rapports

Seuils d’alerte

Estimation des coûts

Prochaines étapes

Ressources