Firebase Summit 2018

I had the opportunity to attend Firebase Summit this year, in Prague, on October 29th. It was not only a mixture of great workshops and sessions, but something that is very important to me: friendly people from all over the world 🙌

You can find all talks online and should definitely watch them if you are using or planning to use Firebase. I’m going to list some of the things announced that I found particularly interesting.

Crashlytics integrated with BigQuery to run deeper analysis on your crash data

You can link BigQuery easily, just select Crashlytics from the left hand side panel (section 1) for your selected project in Firebase and make sure you choose at least 30 days from the top right dropdown (section 2). You’ll see the link to BigQuery in section 3.

At the moment the limits are set as below:
Storage: The BigQuery sandbox has a 10 GB storage limit. Once that limit is reached, data from Firebase will stop exporting to BigQuery.
Dataset expiry: The default expiry for each of your datasets is 60 days. After that time, your data will begin to expire.

Firebase Crashlytics offers integrations with Slack, Jira and PagerDuty

Project Settings (section 1) -> Integrations (section 2)

Here you an choose any (all) of Slack, Jira and PagerDuty (section 3) to get notified when a high impact crash happened.

Performance monitoring dashboard

This gives more insight into issues of your iOS and Android apps. You need to install the Performance Monitoring SDK in the app(s). Once you do this, you can dive into segments of specific traces or network requests in the Firebase console.

Firebase Predictions left BETA and entered general availability

You can just enable Predictions from Firebase Console, nothing needs to be specifically added into your app. By default you get 4 prediction dashboards: churn, not_churn, spend, not_spend and you can add up to 5 more. They’re updated once every 24 hours.

You can also view and analyse raw Predictions data in BigQuery.

Predictions use machine learning to create dynamic user segments based on the predicted behaviour of users in the app and can be used for example to automatically send a notification when users are predicted to stop using the app.

Audience Segmentation

This was existing functionality: the ability to create custom audiences in Firebase Console based on device data, custom events or user properties. However, the audience builder is now enhanced with a few updates: dynamic audience evaluation, audience exclusion and membership duration.

Once you created user segments using Audiences or Predictions, you can use Firebase Cloud Messaging (FCM) to send notifications to latent users to bring them back into your app.

A/B testing entered BETA

You can start experimenting with one of the two (or both) services below:

  • Notifications: experiment with sending messages to engage the right users at the right moment
  • Remote Config: Experiment with app behavior using server-side configuration parameters

Firebase TestLab for iOS left BETA and entered general availability

This is similar to Firebase Android TestLab that was there for a long time. You can now simply run XCTest tests for your iOS app on real devices running in a Google data center. Supported announced versions are iOS 12 and 11. This is a paid service.

Other announcements

  • Various packages for Corporate Support for Firebase will be added by the end of the year. At the moment support for Firebase was just made available for customers already paying for Google Cloud Platform support channel. The Firebase free support will still be there.
  • Incorporated support to deploy to Firebase Hosting with a click of a button from Stackblitz and Glitch web-based IDEs.
  • Remote Config is now integrated with Cloud Functions and Firebase Cloud Messaging. This removes the need for polling the Remote Config, you can just notify the app when a new config is available.
  • ML Kit face recognition can now detect more than 100 points of detail on a face.
  • Local emulators for Cloud Firestore and Realtime Database. They can also be built into CI workflow.

On a personal note, it was a great event, an opportunity to learn, connect with fellow developers around the world, ask questions, see old friends and have fun!

I’m going to end with a collage made from a few photos taken during the summit and my avatar drawn by the talented Virginia Poltrack 🎉 🎉 🎉