Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Run experiments and UX Concepts testing
App Analytics
Study your audience and watch project KPIs - AU, retention, engagement, target events
Improve user experience
Ad Tracking
Track traffic sources and build reports on user acquisition campaigns
Optimize CPI, CPA campaigns
Crash Reporting
Watch app stability
Allow for quick updates on new issues and how to reproduce and fix them
Things to know
Real-time reportsUnlimited data - no quotas on audience and custom eventsAccess to raw data - all data attributes including device IDsFree
One platform to rule them all
Crash ReportingAd TrackingApp Analytics
SDK for all platforms
One platform to rule them all
Behavioral
Behavioral Analytics
User-centric analytics - primary feature for studying and segmentation is user behavior.
Behavioral Segmentation
Apply condition for Custom Event, Install or Launch for separate date ranges
Add up to 5 conditions
Users from given condition can be Included or Excluded from the segment
MAU
Connection Type
WAU
Jailbreak
Session Length
Geo
Platform
Platform Version
Application Version
DAU
Network Provider
Traffic SourceDevice
Screen Resolution
SessionsGender
Age
New Users
Returning Users
Crash Logs
Retention
Rolling Retention
Goals Completions
Behavioral Segmentation
Can be used for all reports to analyze metrics by any dimension
Allows to build almost all queries you can speak out =)
Yandex Music
Build a report to show conversion to second-time purchase
Yandex Music
Show geography and device type of users who had been loyal previously but haven’t launched the app during previous 4 weeks
Build a report to show conversion to second-time purchase
Yandex Music
Show geography and device type of users who had been loyal previously but haven’t launched the app during previous 4 weeks
Compare retention of users who skipped/passed on-boarding
Build a report to show conversion to second-time purchase
Behavioral Segmentation
Behavioral Segmentation
Custom Events | Must Have
Insights
Sign In
Tutorial Complete
Checkout
Purchase
Custom Events | More Useful Data
Not just signal events. Engage event payload to get more breakdowns in reports.
Level CompleteLevelID: 12Score: 74Time: 23
Sign InType: socialProvider: fbUserId: 123456
PurchaseID: com.my.in-appCost: 1.99Currency: USD
Custom Events | Report
Analyze number of devices and number of events with breakdown to event properties values
Custom Events | SegmentationUse properties for segmentation
1. Design and describe your event model. Include event content, conditions and context where it will be sent.2. Implement in application - just single-line reportEvent() call
3. Verify reports to make sure that all is fine
4. Ready for production usage
Custom Events | Steps to start
Users Profiles
Build CRM-like with Custom Events. This allows to explore a single user path and history
Users Profiles
User ProfileID: someuuidName: Dave
1. Add User Profile event. It can be sent once or more. Event should uniquely identify a user.
2. Design event content - which identifiers or data should be passed and stored?
3. Segment desired report to get the path of this user
Hypothesis Testing
A/B and other tests are good way to get new insights, measure new features perception, assess effectiveness of marketing activities.
Hypothesis Testing
1. Prepare hypothesis and determine resulting report you expect. Put appropriate marker events.
2. Correctly distribute the experiment to audience samples
3. Analyze the result
HMarkerExperiment: experimentIDTimestamp: 1234567
Hypothesis Testing
Simple experiment to improve conversion to authorized users
〉Carefully design event model for your application
Summing Up
Take step towards user-centric analytics
〉Develop metrics and scenarios
Related Documents
