Brand Analytics Management: Measuring CLV Across Platforms, Devices and Apps
Download as PPTX, PDF2 likes448 views
The document discusses strategies for managing and monetizing player data in mobile gaming through data analytics systems that ensure flexibility, security, and scalability. It highlights the importance of accurate metrics, access to raw data, and integration with various analytical tools, while showcasing examples of data processing workflows and the use of time series prediction models. Overall, it emphasizes the need to understand user behavior and optimize revenue by analyzing customer acquisition costs and lifetime value across different games.
![Schemaless-ness-ness
time v
1384160400 {“ip”:”135.52.211.23”, “code”:”0”}
1384162200
{“ip”:”45.25.38.156”, “code”:”-1”, “action”:”upload”}
1384164000
{“ip”:”97.12.76.55”, “code”:”99”, “status”:”ok”}
SELECT v[‘ip’] as ip, v [‘code’] as code …
Default (schema-less)
time ip : string code : int action : string
1384160400 135.52.211.23 0 NULL
1384162200
45.25.38.156 -1 upload
1384164000
97.12.76.55 99 NULL
Schema applied ~30% faster
SELECT ip, code …
• Stored “schema-less” as JSON
• Schema can be applied/updated AFTER storage
• Compressed & columnar format
• For higher query performance
• Optimized for time-based filtering
• Quickly scale-up processing power
• WITHOUT reloading/redistributing the idea](https://image.slidesharecdn.com/seattlemgfv92-180605183755/85/Brand-Analytics-Management-Measuring-CLV-Across-Platforms-Devices-and-Apps-14-320.jpg)
Brand Analytics Management: Measuring CLV Across Platforms, Devices and Apps
- 1. BRAND ANALYTICS MANAGEMENT How to unify and monetize player data across platforms, devices and apps
- 3. WHY AM I TALKING TO YOU NOW?
- 4. OUR USE CASE
- 5. Mobile gaming back then: a long time ago in a galaxy far, far away OS STICKINESSDOWNLOADS COUNTRY AVG SESSION TIME MAU /DAU
- 6. Onboarding Minnows OS version RMT AVG Session time KPI eCPM MAU ACPU ACPU ACU ARPPU Crash rate per device Segmentation Conversion rate CPA D28 RK factor LTV PCU
- 7. Transparency Key points for a data management system Flexibility Security Scalability
- 8. TRANSPARENCY
- 9. You made three mistakes • You don’t know how *exactly each metrics are computed • You don’t have access to raw data for reprocessing / debugging • You have no capacity to integrate easily with other tools
- 10. FLEXIBILITY
- 11. Sample game data pipeline Appannie DL & REV Appfigures DL & REV Awesome Ads Ad Revenue Fyber Ad Revenue Chart boost Ad Revenue Apptopia User graph AppAnnie Store views Ooyala Video usage Apptopia DL Rev for Competitors Apptopia KPIs for Competitors Apptopia Retentions for Competitors DOWNLOADS & REVENUE COMPETITORS ESTIMATES RAW EVENTS SOURCE STORE & USAGE METRICS MANUAL IMPORTS Swrve SDK TD Unity SDK Adjust SDK P&L Developer CSV Customer Report Developer PDF Report Customer Extraction Scripts SQL Pre-processing Treasure Data Transform and Export SQL Queries Tableau Dashboards
- 12. Daily ETL Workflow Hardware Usage Processing Export Export Export Import eternal Logs KPI Calculation Process Screen Views Processing Pre Processing Transform & Export SQL Queries Powerd By Workflow Hourly ETL Workflow (Launch Analysis) DL/REV Process KPI Calculation Process Import external Logs Import external Logs Sources Aggregations Export Hourly ETL Workflow (Launch Analysis) DL/REV Estimates Processing Retention Estimates Processing Demographics Estimates Processing App Usage Estimates Processing Appannie DL & REV Appfigures DL & REV Awesome Ads Ad Revenue Fyber Ad Revenue Chart boost Ad Revenue Apptopia User graph AppAnnie Store views Apptopia DL Rev for Competitors Apptopia KPIs for Competitors Apptopia Retentions for Competitors DL & REV SOURCES COMPETITORS ESTIMATES RAW EVENTS SOURCE STORE & USAGE METRICS Swrve SDK TD Unity SDK Adjust SDK Customer Extraction Scripts SQL Pre-processing Treasure Data Transform and Export SQL Queries Tableau Dashboards Ooyala Video usage AppAnnie Store Raitings Apptopia Demography MANUAL IMPORTS AND CUSTOM REPORTS Gameloft KPI Custom Report Flurry KPIS Hot Wheels Custom Report Budge Games KPIS from Tableau Scrabble KPI Custom Report TD Connectors Preprocessing Query External Source RAW Event Table Log Table KPI Table Screen Table Hardware Table KPI Calculation Query Screen View Query Hardware Usage Query Daily Import TD SDKGame Log TableStream KPI Calculation Query Screen View Query Hardware Usage Query KPI Table Screen Table Hardware Table Individual Process for a Game tracked with TD SDK Individual process for a Game not tracked into TD Single game database anatomy
- 13. Workflow Daily ETL Workflow Hardware Usage Processing Export Export Export KPI Calculation Process Screen Views Processing Pre Processing Transform & Export SQL Queries Powerd By Workflow Hourly ETL Workflow (Launch Analysis) DL/REV Process KPI Calculation Process Import external Logs Import external Logs Export Hourly ETL Workflow (Launch Analysis) DL/REV Estimates Processing Retention Estimates Processing Demographics Estimates Processing App Usage Estimates Processing Sources Aggregations Import eternal Logs
- 14. Schemaless-ness-ness time v 1384160400 {“ip”:”135.52.211.23”, “code”:”0”} 1384162200 {“ip”:”45.25.38.156”, “code”:”-1”, “action”:”upload”} 1384164000 {“ip”:”97.12.76.55”, “code”:”99”, “status”:”ok”} SELECT v[‘ip’] as ip, v [‘code’] as code … Default (schema-less) time ip : string code : int action : string 1384160400 135.52.211.23 0 NULL 1384162200 45.25.38.156 -1 upload 1384164000 97.12.76.55 99 NULL Schema applied ~30% faster SELECT ip, code … • Stored “schema-less” as JSON • Schema can be applied/updated AFTER storage • Compressed & columnar format • For higher query performance • Optimized for time-based filtering • Quickly scale-up processing power • WITHOUT reloading/redistributing the idea
- 15. ARIMA time series prediction 0 11 21 32 42 53 1/1/17 2/1/17 3/1/17 4/1/17 5/1/17 6/1/17 7/1/17 8/1/17 9/1/17 10/1/17 11/1/17 12/1/17 1/1/18 2/1/18 ARIMA Prediction Exemple Actual D1 version 1.0 Predicted D1 version 1.0 Actual D1 version 1.1
- 16. SECURITY
- 18. SCALABILITY
- 19. UNIVERSAL LAW IS FOR LACKEYS. CONTEXT IS FOR KINGS.
- 20. Churn prediction Churn Stays >20<20>=0.7<0.7>=10<10>50<=50 <=0.5>0.5 <=100>100 <=10>10 Churn StaysChurn StaysChurn Stays Number of Sessions Number of Sessions HC Earned Victory Rate HC Earned SC earnedVictory Rate
- 21. Sentiment analysis AppStore Reviews Loadings (240) Ads (120) 3 Stars 4 Stars 5 Star Content (30) Video (325) 1 Star 2 Stars
- 22. PART 2 THE BIG PICTURE
- 23. YOU GUYS ALL GOOD ?
- 24. Each game project is (frequently) a silo
- 26. THE BIG PICTURE (Granular) Sarah Connor CPA: 2.14$USD GAME 1 IAP 0$ GAME 2 IAP 1.99$ GAME 1 Ad Rev 0.25$ GAME 2 Ad Rev 0.10$ GAME 3 IAP 0$ GAME 3 Ad Rev 0.30$ GAME 1 LTV 0.25$ GAME 2 LTV 2.09$ GAME 3 LTV 0.30$ ALL GAMES LTV 2.64$
- 27. THE BIG PICTURE Great user quality AVG CPA: 2.14$USD Campaign:A,B,C,.. AVG ALL GAMES IAP REV 2.10$ ALL GAMES IAP REV 0.99$ AVG ALL GAMES AD Rev 1.25$ LTV vs CPA 157% OK user quality AVG CPA: 1.45$USD Campaign:D,E,F,.. Bad user quality AVG CPA: 1.35$USD Campaign:G,H,I,.. X 100k X 100k X 100k ALL GAMES IAP REV 0.79$ AVG ALL GAMES AD Rev 0.55$ AVG ALL GAMES AD Rev 0.25$ LTV vs CPA 106% LTV vs CPA 77%
- 28. Suppose each game is part of a Burger Meal COST 3$ PRICE 3.99$ ROI 0.99$ (25%) COST 0.05$ PRICE 1.29$ ROI 1.24$ (96%) COST 0.10$ PRICE 1.79$ ROI 1.69$ (94%)
- 29. Let’s focus on the highest revenue drivers only? COST 0.05$ PRICE 1.29$ ROI 1.24$ (96%) COST 0.10$ PRICE 1.79$ ROI 1.69$ (94%)
- 30. Let’s focus on the highest revenue drivers only? COST 0.05$ PRICE 1.29$ ROI 1.24$ (96%) COST 0.10$ PRICE 1.79$ ROI 1.69$ (94%)
- 33. EXAMPLES