GA4追踪AI流量
配置与分析全流程指南

# GA4自定义渠道组配置
# 在GA4后台手动配置，以下是配置参数说明

# AI搜索渠道配置
ai_search_channel = {
    'channel_name': 'AI搜索',
    'description': '来自AI搜索引擎的引荐流量',
    'rules': [
        # 规则1: 来源域名匹配AI搜索平台
        {
            'condition': 'source matches regex',
            'pattern': 'chatgpt\.com|perplexity\.ai|kimi\.moonshot\.cn|yiyan\.baidu\.com|phind\.com|poe\.com|you\.com|perplexity\.ai',
            'medium': 'referral'
        },
        # 规则2: UTM参数标记为ai_search
        {
            'condition': 'utm_medium equals',
            'value': 'ai_search',
            'source': 'utm_source'
        }
    ]
}

# 打印配置说明
print('=== AI搜索渠道配置说明 ===')
print(f'渠道名称: {ai_search_channel['channel_name']}')
print(f'描述: {ai_search_channel['description']}')
print(f'\n匹配规则:')
for i, rule in enumerate(ai_search_channel['rules'], 1):
    print(f'  规则{i}:')
    for k, v in rule.items():
        print(f'    {k}: {v}')

print(f'\n包含的AI平台域名:')
ai_domains = [
    ('chatgpt.com', 'ChatGPT Search'),
    ('perplexity.ai', 'Perplexity AI'),
    ('kimi.moonshot.cn', 'Kimi搜索'),
    ('yiyan.baidu.com', '百度文心一言'),
    ('phind.com', 'Phind'),
    ('poe.com', 'Poe'),
    ('you.com', 'You.com'),
]
for domain, name in ai_domains:
    print(f'  {domain:25s} → {name}')

print(f'\n配置位置: GA4后台 → 管理 → 数据收集和修改 → 渠道组')
print(f'配置类型: 创建新渠道组 → 添加AI搜索渠道')

# UTM参数规范与批量生成
import urllib.parse

# UTM参数规范
utm_spec = {
    'utm_source': {
        'description': '来源平台',
        'examples': {
            'chatgpt': 'ChatGPT搜索中的引用',
            'perplexity': 'Perplexity搜索中的引用',
            'kimi': 'Kimi搜索中的引用',
            'zhihu': '知乎文章中的链接',
            'medium': 'Medium文章中的链接',
            'github': 'GitHub README中的链接',
            'baijiahao': '百家号文章中的链接'
        }
    },
    'utm_medium': {
        'description': '媒介类型',
        'standard_value': 'ai_search',
        'note': '所有AI搜索相关链接统一使用ai_search'
    },
    'utm_campaign': {
        'description': '活动名称',
        'naming_convention': 'geo_{type}_{period}',
        'examples': ['geo_faq_2026q3', 'geo_techblog_2026aug', 'geo_industryreport_2026']
    },
    'utm_content': {
        'description': '具体内容标识',
        'naming_convention': '{content_type}_{keyword}',
        'examples': ['faq_crm_features', 'blog_ai_prediction', 'report_crm_market_2026']
    }
}

# 批量生成UTM链接
def generate_utm_urls(base_url, sources, medium='ai_search', campaign='geo_content_2026q3', contents=None):
    '''批量生成带UTM参数的链接'''
    urls = []
    for source in sources:
        if contents:
            for content in contents:
                params = {
                    'utm_source': source,
                    'utm_medium': medium,
                    'utm_campaign': campaign,
                    'utm_content': content
                }
                utm_url = base_url + '?' + urllib.parse.urlencode(params)
                urls.append({'source': source, 'content': content, 'url': utm_url})
        else:
            params = {
                'utm_source': source,
                'utm_medium': medium,
                'utm_campaign': campaign
            }
            utm_url = base_url + '?' + urllib.parse.urlencode(params)
            urls.append({'source': source, 'content': '', 'url': utm_url})
    return urls

# 示例：为产品页面生成不同平台的UTM链接
base_url = 'https://your-domain.com/products/crm-pro'
sources = ['zhihu', 'medium', 'github', 'baijiahao']
contents = ['faq_crm_features', 'blog_ai_prediction', 'tutorial_crm_setup']

utm_urls = generate_utm_urls(base_url, sources, contents=contents)
print(f'共生成{len(utm_urls)}条UTM链接:\n')
for item in utm_urls[:6]:  # 只展示前6条
    print(f'[{item['source']}/{item['content']}]')
    print(f'  {item['url']}')
    print()

# GA4 API查询AI搜索流量数据
# 需要安装: pip install google-analytics-data

from google.analytics.data_v1beta import BetaAnalyticsDataClient
from google.analytics.data_v1beta.types import (
    RunReportRequest, DateRange, Dimension, Metric, 
    FilterExpression, Filter
)

def query_ai_search_traffic(property_id, credentials_path, start_date, end_date):
    '''查询AI搜索引荐流量数据'''
    from google.oauth2 import service_account
    
    credentials = service_account.Credentials.from_service_account_file(
        credentials_path,
        scopes=['https://www.googleapis.com/auth/analytics.readonly']
    )
    client = BetaAnalyticsDataClient(credentials=credentials)
    
    # AI搜索来源域名正则
    ai_domains = 'chatgpt.com|perplexity.ai|kimi.moonshot.cn|yiyan.baidu.com|phind.com'
    
    request = RunReportRequest(
        property=f'properties/{property_id}',
        date_ranges=[DateRange(start_date=start_date, end_date=end_date)],
        dimensions=[
            Dimension(name='sessionSource'),      # 引荐来源
            Dimension(name='sessionMedium'),      # 媒介
            Dimension(name='landingPage'),         # 着陆页
            Dimension(name='deviceCategory'),      # 设备类型
        ],
        metrics=[
            Metric(name='sessions'),              # 会话数
            Metric(name='totalUsers'),            # 用户数
            Metric(name='screenPageViews'),       # 页面浏览量
            Metric(name='averageSessionDuration'), # 平均会话时长
            Metric(name='bounceRate'),            # 跳出率
            Metric(name='conversions'),           # 转化数
        ],
        dimension_filter=FilterExpression(
            filter=Filter(
                field_name='sessionSource',
                string_filter=Filter.StringFilter(
                    match_type=Filter.StringFilter.MatchType.PARTIAL_REGEXP,
                    value=ai_domains
                )
            )
        )
    )
    
    response = client.run_report(request)
    
    # 解析结果
    results = []
    for row in response.rows:
        result = {}
        for i, dim in enumerate(row.dimension_values):
            result[response.dimension_headers[i].name] = dim.value
        for i, metric in enumerate(row.metric_values):
            result[response.metric_headers[i].name] = metric.value
        results.append(result)
    
    return results

# 使用示例（需要实际API凭证）
# results = query_ai_search_traffic(
#     property_id='your_property_id',
#     credentials_path='service-account.json',
#     start_date='2026-08-01',
#     end_date='2026-08-31'
# )
# for r in results:
#     print(f'来源: {r.get('sessionSource', 'N/A')}')
#     print(f'会话数: {r.get('sessions', 'N/A')}')
#     print(f'跳出率: {r.get('bounceRate', 'N/A')}')
#     print(f'转化数: {r.get('conversions', 'N/A')}')
#     print('-' * 40)

print('GA4 API查询脚本已准备就绪')
print('使用前需要: 1)创建GCP服务账号 2)启用GA4 Data API 3)下载JSON凭证')
print('文档: https://developers.google.com/analytics/devguides/reporting/data/v1')