📡 Idea Radar 每天 8:00 / 18:00 推送 · 18:00 生成精选
reddit · SideProject · 46 分 ← 返回列表

How we built an API to detect bot-inflated domain traffic (headless browser analysis + search heuristics)

Hey everyone, If you build tools in ad-tech, affiliate marketing, programmatic SEO, or domain acquisition, you’ve probably run into this problem: reported website traffic is trivial to fake. Traditional SEO platforms and estimators calculate traffic by looking at Google keyword positions and multiplying them by static CTR models. Spammers take advantage of this by ranking expired domains for high-volume junk keywords, routing residential proxy botnets through search engines, or syndicat

五维评分

付费意愿
48
需求真实度
42
个人可落地
55
竞争空间
30
新鲜度
52

AI 分析

痛点:网站流量数据极易造假,传统SEO工具按关键词排名×固定CTR估算会把零人类流量的站报成8万/月,导致买域名、投广告、做外联时把钱浪费在“幽灵曝光”上。

目标用户:域名投资人/域名经纪、联盟营销与广告技术从业者、程序化SEO团队、广告库存买家与外联获客团队。

证据:reported website traffic is trivial to fake. / To standard SEO tools, the site looks like it’s pulling 80,000 visits/month. In reality, actual human traffic is practically zero. / relying on raw traffic estimates often means burning budget on ghost impressions.

变现:按查询量计费的REST API订阅(免费额度+阶梯定价),批量域名扫描包月套餐,企业版白标/私有部署与告警监控。

MVP:2周内做一个单域名真实性评分API:对一个域名做少量关键词抽样,用Playwright无头浏览器+住宅代理抓取真实搜索点击与落地表现,叠加“权重-流量均衡”“搜索足迹”“跳转链”“流量来源结构”四条启发式规则,输出0-100分与理由,配一个极简查询页和CSV批量上传。

竞品:SimilarWeb、Ahrefs、Semrush、Fraudlogix、CHEQ、Pixalate、ipinfo/Spur

摘要:帖主介绍了自建的Traffic Authenticity Index及其REST API,用无头Node.js验证worker加搜索足迹分析,识别靠过期域名、住宅代理机器人、pop-under网络刷出的假流量。其称传统SEO工具会把实际人类访问近乎为零的站估算成8万访问/月,让买域名或投广告的预算浪费在幽灵曝光上。痛点真实但属自建产品自述,无第三方付费或收入佐证,且SimilarWeb、CHEQ等成熟竞品密集,独立开发者切入需靠细分场景与更低价格。