答案引擎优化 (AEO) 是 SEO 机构不断增长的服务线，但大多数排名跟踪工具不能很好地跟踪 AI 概述引用或为其收取高价。使用搜索 API 构建自定义 AEO 仪表板使代理机构能够完全控制数据、每个客户的视图，并使每个关键字检查的成本可预测为 0.005 美元。本教程构建了一个多客户端 AEO 仪表板后端，用于跟踪引文存在、语音份额和趋势数据，然后输出可用于任何前端框架的 JSON。
前置条件
- 已安装 Python 3.9+
- 请求已安装库
- 来自 scavio.dev 的 Scavio API 密钥
- 要监控的客户端关键字列表
操作指南
步骤 1: 定义多客户端数据结构
设置将每个客户端映射到其域和目标关键字的配置。这是仪表板的输入。
Python
clients = {
'client_a': {
'domain': 'clienta.com',
'name': 'Client A - SaaS CRM',
'keywords': [
'best crm for small business',
'crm software comparison 2026',
'what is a crm system',
'crm vs spreadsheet'
]
},
'client_b': {
'domain': 'clientb.io',
'name': 'Client B - Project Management',
'keywords': [
'best project management tools 2026',
'how to manage remote teams',
'project management software comparison',
'agile vs waterfall 2026'
]
}
}
total_keywords = sum(len(c['keywords']) for c in clients.values())
print(f'{len(clients)} clients, {total_keywords} total keywords')
print(f'Monthly cost (daily checks): ${total_keywords * 30 * 0.005:.2f}')步骤 2: 构建AEO数据采集功能
对于每个关键字，检查 SERP 中的 AI 概述是否存在并提取引文数据。记录客户端域是否被引用。
Python
import requests, os, time
API_KEY = os.environ['SCAVIO_API_KEY']
def collect_aeo_data(keyword: str, client_domain: str) -> dict:
resp = requests.post('https://api.scavio.dev/api/v1/search',
headers={'x-api-key': API_KEY, 'Content-Type': 'application/json'},
json={'query': keyword, 'country_code': 'us'})
data = resp.json()
aio = data.get('ai_overview', {})
has_overview = bool(aio and aio.get('text'))
citations = aio.get('citations', []) if has_overview else []
cited_domains = [c.get('domain', '') for c in citations]
# Also check organic position
organic_position = None
for r in data.get('organic_results', []):
if client_domain in r.get('link', ''):
organic_position = r['position']
break
return {
'keyword': keyword,
'has_ai_overview': has_overview,
'total_citations': len(citations),
'client_cited': any(client_domain in d for d in cited_domains),
'citation_position': next(
(i + 1 for i, d in enumerate(cited_domains) if client_domain in d), None),
'organic_position': organic_position,
'competitor_domains': [d for d in cited_domains if client_domain not in d]
}步骤 3: 生成每个客户端的仪表板数据
将关键字级数据聚合为客户级指标：引用率、平均引用位置、有机重叠和主要竞争对手领域。
Python
from collections import Counter
def generate_client_dashboard(client_id: str, config: dict) -> dict:
domain = config['domain']
results = []
competitor_counter = Counter()
for kw in config['keywords']:
data = collect_aeo_data(kw, domain)
results.append(data)
for d in data['competitor_domains']:
competitor_counter[d] += 1
time.sleep(0.3)
total = len(results)
with_aio = sum(1 for r in results if r['has_ai_overview'])
cited = sum(1 for r in results if r['client_cited'])
avg_cite_pos = 0
cite_positions = [r['citation_position'] for r in results if r['citation_position']]
if cite_positions:
avg_cite_pos = sum(cite_positions) / len(cite_positions)
return {
'client_id': client_id,
'client_name': config['name'],
'domain': domain,
'total_keywords': total,
'keywords_with_aio': with_aio,
'aio_rate': round(with_aio / max(total, 1) * 100, 1),
'citations': cited,
'citation_rate': round(cited / max(with_aio, 1) * 100, 1),
'avg_citation_position': round(avg_cite_pos, 1),
'top_competitors': competitor_counter.most_common(5),
'keyword_details': results
}步骤 4: 构建完整的机构仪表板输出
为所有客户端运行仪表板并输出可供前端使用的 JSON 结构。包括每个客户的成本跟踪。
Python
import json
from datetime import date
def build_agency_dashboard(clients: dict) -> dict:
dashboard = {
'generated': date.today().isoformat(),
'clients': [],
'summary': {}
}
total_credits = 0
for client_id, config in clients.items():
print(f'Processing {config["name"]}...')
client_data = generate_client_dashboard(client_id, config)
dashboard['clients'].append(client_data)
total_credits += len(config['keywords'])
# Agency summary
all_clients = dashboard['clients']
dashboard['summary'] = {
'total_clients': len(all_clients),
'total_keywords': sum(c['total_keywords'] for c in all_clients),
'avg_citation_rate': round(
sum(c['citation_rate'] for c in all_clients) / max(len(all_clients), 1), 1),
'total_credits': total_credits,
'cost': f'${total_credits * 0.005:.2f}'
}
with open(f'aeo_dashboard_{date.today()}.json', 'w') as f:
json.dump(dashboard, f, indent=2)
return dashboard
dashboard = build_agency_dashboard(clients)
for c in dashboard['clients']:
print(f'{c["client_name"]}: {c["citation_rate"]}% citation rate, {c["citations"]}/{c["keywords_with_aio"]} cited')Python 示例
Python
import os, requests, time, json
from datetime import date
from collections import Counter
API_KEY = os.environ['SCAVIO_API_KEY']
def check_aeo(keyword, domain):
resp = requests.post('https://api.scavio.dev/api/v1/search',
headers={'x-api-key': API_KEY, 'Content-Type': 'application/json'},
json={'query': keyword, 'country_code': 'us'})
aio = resp.json().get('ai_overview', {})
citations = aio.get('citations', []) if aio.get('text') else []
domains = [c.get('domain', '') for c in citations]
return {'keyword': keyword, 'has_aio': bool(aio.get('text')),
'cited': any(domain in d for d in domains), 'domains': domains}
def client_report(name, domain, keywords):
results = [check_aeo(kw, domain) for kw in keywords]
cited = sum(1 for r in results if r['cited'])
with_aio = sum(1 for r in results if r['has_aio'])
print(f'{name}: {cited}/{with_aio} cited ({len(keywords)} keywords, ${len(keywords)*0.005:.3f})')
client_report('Demo Client', 'example.com', ['best crm 2026', 'crm comparison'])JavaScript 示例
JavaScript
const API_KEY = process.env.SCAVIO_API_KEY;
async function checkAeo(keyword, domain) {
const resp = await fetch('https://api.scavio.dev/api/v1/search', {
method: 'POST',
headers: { 'x-api-key': API_KEY, 'Content-Type': 'application/json' },
body: JSON.stringify({ query: keyword, country_code: 'us' })
});
const data = await resp.json();
const aio = data.ai_overview || {};
const domains = (aio.citations || []).map(c => c.domain || '');
return { keyword, hasAio: Boolean(aio.text), cited: domains.some(d => d.includes(domain)) };
}
async function main() {
const keywords = ['best crm 2026', 'crm comparison'];
let cited = 0;
for (const kw of keywords) {
const r = await checkAeo(kw, 'example.com');
if (r.cited) cited++;
console.log(`${r.hasAio ? '+' : '-'}AIO ${r.cited ? '+CITED' : ''} ${kw}`);
}
console.log(`Citation rate: ${(cited/keywords.length*100).toFixed(0)}%`);
}
main().catch(console.error);预期输出
JSON
2 clients, 8 total keywords
Monthly cost (daily checks): $1.20
Processing Client A - SaaS CRM...
Processing Client B - Project Management...
Client A - SaaS CRM: 50.0% citation rate, 1/2 cited
Client B - Project Management: 33.3% citation rate, 1/3 cited
Agency summary:
Total clients: 2
Total keywords: 8
Avg citation rate: 41.7%
Cost: $0.04