At the moment of creating this library (June 2016) there is no watchos support for the Google Analytics library. To still be able to send your analytics to Google or any other analytics service this library provides an interface that can be used by watchos apps. It uses the WatchConnectivity framework under the hood to first send analytics requests from the Watch app to the iPhone app. Then it's send automatically from the iPhone app to the analytics service.

To run the example project, clone the repo, and run pod install from the Example directory first.

Since this library is using the WatchConnectivity framework it will only work on iOS 9+ and watchos 2+. Though due to the use of #availablity it is source compatible with iOS 8.

AnalyticsWatchAdapter is available through CocoaPods. To install it, simply add the following line to your Podfile:

pod "AnalyticsWatchAdapter"

Include this both in your iOS target and WatchKit extension target.

Tracking analytics is done through the Analytics protocol. Wether it's used from the iOS or watchos target, the usuage is the same though setup differs (see sections below). An instance of this protocol is obtained through the global analyticsInstance singleton instance.

To track a page/screen view use:

analyticsInstance?.trackPageViewAsString("Welcome screen")

You can also track a category:

analyticsInstance?.trackCategory("store", action: "purchase")

For more options, see the Analytics protocol.

For this library to work you must provide an implementation of the Analytics protocol in your iOS target.

This library does not include a default implementation of the Analytics protocol. However, an implementation using Google Analytics is included in the sample code in GoogleAnalytics.swift. For that to work, copy this file into your project and add a pod dependency to 'GoogleAnalytics'. (Because the GoogleAnalytics library contains static libraries it cannot be a transitive dependency of this library, see CocoaPods/CocoaPods#2926).

You can provide your own implementation of Analytics. Make sure to set the global analyticsInstance to refer to an instance of your implementation.

The previous step is enough if you only want send analytic events from your iOS target. If however you also want to send events from your watchos target (which is actually the prupose of this library) you need to set the AnalyticsWatchReceiver as delegate of your WCSession. To learn how to setup a WCSession see the documentation at https://developer.apple.com/library/ios/documentation/WatchConnectivity/Reference/WCSession_class/.

You can either use the AnalyticsWatchReceiver directly as delegate ...

if WCSession.isSupported() {
    let session = WCSession.defaultSession()
    session.delegate = AnalyticsWatchReceiver.instance
    session.activateSession()
}

... or if you also need to receive other events from the Watch App implement a proxy that forwards the session:didReceiveUserInfo: method to AnalyticsWatchReceiver. You can use the isAnalyticUserInfo method to check if something should be handled by the AnalyticsWatchReceiver:

func session(session: WCSession, didReceiveUserInfo userInfo: [String : AnyObject]) {
    if AnalyticsWatchReceiver.isAnalyticUserInfo(userInfo) {
        AnalyticsWatchReceiver.instance.session(session, didReceiveUserInfo: userInfo)
    }
}

In your watchos extension target you need to start the AnalyticsWatchSender with the WCSession:

let session = WCSession.defaultSession()
session.delegate = WatchModel.instance
session.activateSession()
AnalyticsWatchSender.start(session)

This will set an implementation of the Analytics protocol to the global analyticsInstance that sends the events to the phone. A good place for this would be inside awakeWithContext of your initial Interface Controller.

After that you can send analytic events from the watch app just like you would from you iOS app:

analyticsInstance?.trackPageViewAsString("Welcome screen in Watch App")

Lammert Westerhoff, [email protected]

AnalyticsWatchAdapter is available under the MIT license. See the LICENSE file for more info.