Google API：我自己的 PPC 解决方案应使用哪个 API？
我将开发一个网站，允许广告商发布他们的产品/服务。每个产品/服务都会有一个返回广告商网站的链接。我计划实施按点击付费的商业模式。
我是否可以使用 Google API 来跟踪广告商列表的点击次数？我对已经捕获所有类型指标、检测点击欺诈等的解决方案感兴趣。我还想向广告商提供有关其广告的报告。
我认为一种可能的解决方案是使用 Google Analytics。由于我将在数据库中存储返回广告商网站的链接，因此我想我可以使用 Analytics 中的“退出”信息来确定有多少用户点击了广告商的链接。这是一个好的解决方案还是有更好的解决方案？
I am going to develop a site that allow advertisers to post their products/services. Each product/service will have a link back to the advertiser's site. I plan to implement a pay-per-click business model.
Is there a Google API I can use to keep track of the clicks to the advertiser's listings? I'm interested in a solution that already captures all sort of metrics, detect click frauds, etc. I also want to provide advertisers reports about their ads.
One possible solution, I think, is to use Google Analytics. Since I will store in my database the link back to the advertiser's site, I suppose I could use the "exit" information from Analytics to determine how many users clicked on the advertiser's link. Is this a good solution or is there a better one?
如果你对这篇内容有疑问，欢迎到本站社区发帖提问 参与讨论，获取更多帮助，或者扫码二维码加入 Web 技术交流群。
绑定邮箱获取回复消息
由于您还没有绑定你的真实邮箱，如果其他用户或者作者回复了您的评论，将不能在第一时间通知您！
发布评论
评论（1）
您可以使用 Google Analytics API 的 trackEvent 功能来跟踪事件：
但是，这只能为您提供简单的信息。尝试依赖 javascript api 解决方案的问题在于，并非所有用户都会启用 javascript。对于付费服务来说，这不够准确。
如果您希望客户为此服务付费，您将需要实施服务器端解决方案。这意味着每个外部链接都将被重定向脚本的链接替换。该脚本会将您转发到实际站点，但在执行此操作时，它会记录有关用户的点击和信息，例如他们的 IP 地址等。
有了这些信息，您就可以使用它来检测模式，例如来自 IP 的多次点击在短时间内，或者每天点击一次，或者您决定的任何模式。
可能值得研究一下，看看是否有现成的解决方案可以消除必须了解点击欺诈检测的麻烦。
You could track the event using the trackEvent feature of the Google Analytics API:
However this would only give you simple information. The problem with trying to rely on a javascript api solution is that not all users will have javascript enabled. This is not enough accuracy for a paid service.
If you expect customers to pay for this service you are going to need to implement a server side solution. This means that every external link would be replaced by a link to a redirect script. The script would forward you on to the actual site but as it was doing so it would register the click and information about the user such as their IP address etc.
With this information you can then use it to detect patterns like multiple clicks from an IP in a short period, or things like a click every day, or whatever patterns you decide.
It might be worth looking into this to see if there is an off the shelf solution which will take away the hassle from having to learn about click fraud detection.