How to obtain the existing Google Analytic clientId on OS X and/or Windows for use with Google Measurement Protocol

Question

We are extending our analytic coverage to include actions on our Desktop apps for Windows and Mac OSX.

The Google Measurement Protocol is simple enough to use: https://developers.google.com/analytics/devguides/collection/protocol/v1/

Is there a way that we can extract the existing Google client-id, that is, the id generated by Google and generally stored in a cookie on the computer that uniquely and anonymously identifies the client?

For clarity - you would typically use Javascript to extract it in a web-page as follows; https://developers.google.com/analytics/devguides/collection/analyticsjs/cookies-user-id?hl=en#getting_the_client_id_from_the_cookie

How do you do the same using a language such as Swift or Objective-C on OS X, or MFC/C++ on Windows.

Answer 1

The client id is generated by the Javascript tracking code (or the SDK respectively). So if you use the measurement protocol there is no pre-existing client id that you could extract.

The good news is that you can generate your own. Google Analytics isn't critical w/r/t what it accepts as client id, any string will do - however to serve it's intended purpose it must be unique, so a UUID is recommended.

Answer 2

The best way to do this would be to use the User ID feature of Google Analytics. This feature lets you send your own ID instead of using the randomly generated Client ID (that is stored in a cookie).

So if the user visits your website and identifies themselves (e.g. logs in), then you would sent their "customer ID" using the User ID feature. Then, if that user goes on to install your desktop application and also identifies themselves (by logging in again), then you would use the measurement protocol to send the same "customer ID". You would do this by using the &uid measurement protocol parameter. Google Analytics will then identify this as the same user an you will basically get x-device behaviour in GA. This will allow you to figure out that campaign A drove a website visit as well as an eventual desktop app download.