Comment créer des rapports automatisés de visibilité GEO

import os, requests, json
from datetime import datetime
from collections import defaultdict

API_KEY = os.environ['SCAVIO_API_KEY']
SH = {'x-api-key': API_KEY, 'Content-Type': 'application/json'}

BRAND = 'scavio.dev'
COMPETITORS = ['tavily.com', 'serpapi.com']
KEYWORDS = [
    'best search api for ai agents',
    'web search api comparison',
    'mcp search tool setup',
    'serp api free tier',
    'search api for rag pipeline',
]

def geo_score_keyword(keyword, domain):
    data = requests.post('https://api.scavio.dev/api/v1/search',
        headers=SH, json={'query': keyword, 'country_code': 'us'}, timeout=10).json()
    ai = data.get('ai_overview', data.get('answer_box', {}))
    organic = data.get('organic_results', [])
    featured = data.get('featured_snippet', {})
    paa = data.get('people_also_ask', [])
    domain_l = domain.lower()
    # Score components (0-100 each)
    ai_score = 30 if (ai and domain_l in json.dumps(ai).lower()) else 0
    org_pos = next((i+1 for i, r in enumerate(organic) if domain_l in r.get('link', '').lower()), None)
    org_score = max(0, 30 - (org_pos - 1) * 3) if org_pos else 0
    feat_score = 20 if (featured and domain_l in json.dumps(featured).lower()) else 0
    paa_score = 20 if any(domain_l in json.dumps(q).lower() for q in paa) else 0
    total = ai_score + org_score + feat_score + paa_score
    return {'keyword': keyword, 'total': total, 'ai': ai_score, 'organic': org_score,
            'featured': feat_score, 'paa': paa_score, 'organic_pos': org_pos}

print(f'GEO Visibility Scan for {BRAND}\n')
scores = [geo_score_keyword(kw, BRAND) for kw in KEYWORDS]
for s in scores:
    pos = f'#{s["organic_pos"]}' if s['organic_pos'] else '-'
    print(f'  {s["keyword"][:40]:40} | GEO: {s["total"]:3} | Pos: {pos}')
avg = sum(s['total'] for s in scores) / len(scores)
print(f'\nAverage GEO Score: {avg:.0f}/100 | Cost: ${len(KEYWORDS) * 0.005:.3f}')

def competitive_geo(keywords, brand, competitors):
    all_domains = [brand] + competitors
    domain_scores = defaultdict(list)
    for kw in keywords:
        for domain in all_domains:
            score = geo_score_keyword(kw, domain)
            domain_scores[domain].append(score)
    print(f'\n=== GEO Competitive Analysis ===')
    rankings = []
    for domain, keyword_scores in domain_scores.items():
        avg = sum(s['total'] for s in keyword_scores) / len(keyword_scores)
        ai_pct = sum(1 for s in keyword_scores if s['ai'] > 0) / len(keyword_scores) * 100
        rankings.append({'domain': domain, 'avg_geo': avg, 'ai_citation_pct': ai_pct})
    rankings.sort(key=lambda x: x['avg_geo'], reverse=True)
    for i, r in enumerate(rankings, 1):
        marker = ' <-- you' if r['domain'] == brand else ''
        print(f'  {i}. {r["domain"]:25} | GEO: {r["avg_geo"]:5.1f} | AI cited: {r["ai_citation_pct"]:.0f}%{marker}')
    return rankings

rankings = competitive_geo(KEYWORDS, BRAND, COMPETITORS)
print(f'\nTotal cost: ${len(KEYWORDS) * (1 + len(COMPETITORS)) * 0.005:.3f}')

def weekly_geo_report(brand, scores, rankings):
    print(f'\n{"=" * 60}')
    print(f'  WEEKLY GEO VISIBILITY REPORT')
    print(f'  Brand: {brand} | Date: {datetime.now().strftime("%Y-%m-%d")}')
    print(f'{"=" * 60}')
    avg_geo = sum(s['total'] for s in scores) / len(scores)
    ai_cited = sum(1 for s in scores if s['ai'] > 0)
    print(f'\n  Overall GEO Score: {avg_geo:.0f}/100')
    print(f'  AI Citations: {ai_cited}/{len(scores)} keywords')
    print(f'  Competitive Rank: {next((i+1 for i, r in enumerate(rankings) if r["domain"] == brand), "?")} of {len(rankings)}')
    # Breakdown
    print(f'\n  Score Breakdown:')
    print(f'    AI Overview:  {sum(s["ai"] for s in scores)/len(scores):.0f}/30')
    print(f'    Organic:      {sum(s["organic"] for s in scores)/len(scores):.0f}/30')
    print(f'    Featured:     {sum(s["featured"] for s in scores)/len(scores):.0f}/20')
    print(f'    People Ask:   {sum(s["paa"] for s in scores)/len(scores):.0f}/20')
    # Opportunities
    weak = [s for s in scores if s['total'] < 30]
    if weak:
        print(f'\n  Improvement Opportunities:')
        for s in weak:
            print(f'    - {s["keyword"][:40]} (GEO: {s["total"]})')
    print(f'\n  Weekly cost: ${len(KEYWORDS) * (1 + len(COMPETITORS)) * 0.005 * 7:.2f}')

weekly_geo_report(BRAND, scores, rankings)