La génération de leads locaux à grande échelle nécessite de rechercher des entreprises par type et par emplacement, d'extraire les informations de contact et de qualifier les leads. Ce tutoriel construit un pipeline qui recherche les entreprises locales à l'aide des données SERP, extrait les sites web, les e-mails et les numéros de téléphone des résultats, et note chaque lead. Une recherche coûte 0,005 $, et un scan typique ville-catégorie de 20 requêtes coûte 0,10 $.
Prérequis
- Python 3.9+ installé
- bibliothèque requests installée
- Une clé API Scavio depuis scavio.dev
Parcours
Étape 1: Définir les emplacements cibles et les catégories d'entreprises
Configurez les villes et les types d'entreprises à rechercher. Chaque combinaison devient une requête de recherche.
Python
locations = [
'Austin TX', 'Denver CO', 'Portland OR',
'Nashville TN', 'Raleigh NC'
]
categories = [
'plumber', 'electrician', 'HVAC contractor',
'roofing company'
]
total_queries = len(locations) * len(categories)
cost = total_queries * 0.005
print(f'{len(locations)} cities x {len(categories)} categories = {total_queries} queries')
print(f'Estimated cost: ${cost:.2f}')Étape 2: Rechercher des entreprises locales
Exécutez des requêtes de recherche pour chaque paire emplacement-catégorie et collectez les listes d'entreprises à partir des résultats SERP.
Python
import requests, os, time, re
SCAVIO_KEY = os.environ['SCAVIO_API_KEY']
def search_local(category: str, location: str) -> list:
query = f'{category} in {location}'
resp = requests.post('https://api.scavio.dev/api/v1/search',
headers={'x-api-key': SCAVIO_KEY, 'Content-Type': 'application/json'},
json={'query': query, 'country_code': 'us', 'num_results': 10})
resp.raise_for_status()
results = resp.json().get('organic_results', [])
leads = []
for r in results:
domain = r['link'].split('/')[2] if '/' in r['link'] else ''
# Skip aggregator sites
if any(agg in domain for agg in ['yelp.com', 'yellowpages.com',
'homeadvisor.com', 'angi.com', 'thumbtack.com', 'bbb.org']):
continue
leads.append({
'business_name': r['title'].split(' - ')[0].split(' | ')[0].strip(),
'website': r['link'],
'domain': domain,
'snippet': r.get('snippet', ''),
'category': category,
'location': location
})
return leads
# Test one query
leads = search_local('plumber', 'Austin TX')
print(f'Found {len(leads)} leads for plumber in Austin TX')
for l in leads[:3]:
print(f' {l["business_name"]}: {l["domain"]}')Étape 3: Extraire les informations de contact des extraits
Analysez les numéros de téléphone, les e-mails et les adresses des extraits SERP. De nombreuses listes d'entreprises locales incluent directement les informations de contact.
Python
def extract_contacts(lead: dict) -> dict:
text = lead.get('snippet', '') + ' ' + lead.get('business_name', '')
# Phone numbers
phones = re.findall(r'\(?\d{3}\)?[-.\s]?\d{3}[-.\s]?\d{4}', text)
# Email patterns
emails = re.findall(r'[\w.+-]+@[\w-]+\.[\w.-]+', text)
# Rating patterns (e.g., "4.8 stars", "4.8/5")
ratings = re.findall(r'(\d\.\d)\s*(?:stars?|/5|rating)', text.lower())
lead['phone'] = phones[0] if phones else ''
lead['email'] = emails[0] if emails else ''
lead['rating'] = float(ratings[0]) if ratings else 0.0
return lead
# Enrich existing leads
enriched = [extract_contacts(l) for l in leads]
for l in enriched[:3]:
print(f' {l["business_name"]}: phone={l["phone"] or "N/A"}, '
f'rating={l["rating"] or "N/A"}')Étape 4: Exécuter le pipeline complet et exporter les résultats
Scannez toutes les combinaisons emplacement-catégorie, enrichissez les leads, dédupliquez par domaine et exportez en CSV.
Python
import csv
def run_pipeline(locations: list, categories: list) -> list:
all_leads = []
seen_domains = set()
for location in locations:
for category in categories:
leads = search_local(category, location)
for lead in leads:
if lead['domain'] not in seen_domains:
seen_domains.add(lead['domain'])
all_leads.append(extract_contacts(lead))
time.sleep(0.3)
return all_leads
leads = run_pipeline(locations[:2], categories[:2]) # small test
print(f'Total unique leads: {len(leads)}')
# Export to CSV
if leads:
with open('local_leads.csv', 'w', newline='') as f:
writer = csv.DictWriter(f, fieldnames=['business_name', 'domain',
'phone', 'email', 'rating', 'category', 'location', 'website'])
writer.writeheader()
writer.writerows(leads)
print(f'Exported to local_leads.csv')
queries_used = len(locations[:2]) * len(categories[:2])
print(f'Cost: {queries_used} queries = ${queries_used * 0.005:.3f}')Exemple Python
Python
import requests, os, re, time, csv
SCAVIO_KEY = os.environ['SCAVIO_API_KEY']
def find_leads(category, location):
resp = requests.post('https://api.scavio.dev/api/v1/search',
headers={'x-api-key': SCAVIO_KEY, 'Content-Type': 'application/json'},
json={'query': f'{category} in {location}', 'country_code': 'us', 'num_results': 10})
leads = []
for r in resp.json().get('organic_results', []):
domain = r['link'].split('/')[2] if '/' in r['link'] else ''
if not any(a in domain for a in ['yelp.com', 'yellowpages.com', 'angi.com']):
phones = re.findall(r'\(?\d{3}\)?[-.\s]?\d{3}[-.\s]?\d{4}', r.get('snippet', ''))
leads.append({'name': r['title'].split(' - ')[0], 'domain': domain,
'phone': phones[0] if phones else ''})
return leads
leads = find_leads('plumber', 'Austin TX')
for l in leads[:5]:
print(f'{l["name"]}: {l["domain"]} {l["phone"]}')Exemple JavaScript
JavaScript
const SCAVIO_KEY = process.env.SCAVIO_API_KEY;
async function findLeads(category, location) {
const resp = await fetch('https://api.scavio.dev/api/v1/search', {
method: 'POST',
headers: { 'x-api-key': SCAVIO_KEY, 'Content-Type': 'application/json' },
body: JSON.stringify({ query: `${category} in ${location}`, country_code: 'us', num_results: 10 })
});
const data = await resp.json();
return (data.organic_results || []).filter(r => {
const domain = new URL(r.link).hostname;
return !['yelp.com', 'yellowpages.com', 'angi.com'].some(a => domain.includes(a));
}).map(r => ({
name: r.title.split(' - ')[0],
link: r.link,
phone: (r.snippet || '').match(/\(?\d{3}\)?[-.\s]?\d{3}[-.\s]?\d{4}/)?.[0] || ''
}));
}
findLeads('plumber', 'Austin TX').then(leads => leads.forEach(l => console.log(`${l.name}: ${l.phone}`)));Sortie attendue
JSON
5 cities x 4 categories = 20 queries
Estimated cost: $0.10
Found 7 leads for plumber in Austin TX
Radiant Plumbing: radiantplumbing.com
ABC Home Services: abchomeservices.com
Mr Rooter Plumbing: mrrooter.com
Total unique leads: 24
Exported to local_leads.csv
Cost: 4 queries = $0.020