Plan the complete analytics infrastructure for a mobile game including event taxonomy, data pipeline architecture, KPI dashboards, and the tooling stack needed to make data-driven decisions across product, marketing, and operations teams.
## CONTEXT Data-driven decision-making is the single greatest predictor of mobile game success, yet most studios launch with inadequate analytics infrastructure that leaves them flying blind during the critical early months. A properly instrumented game generates 50-100+ event types per user per session, creating datasets that can answer virtually any product, marketing, or business question when properly collected, stored, and analyzed. Studios that invest in analytics infrastructure before launch rather than retrofitting after discover issues 2-3x faster, iterate 4x more efficiently on features, and make monetization decisions with 80% higher confidence. The analytics stack has also grown more complex with privacy regulations (GDPR, CCPA, ATT), cross-platform play requirements, and the increasing sophistication of machine-learning models that require clean, comprehensive data. The cost of getting analytics wrong is measured in months of wasted development on features that do not impact KPIs, marketing spend allocated to underperforming channels, and game-economy problems that are detected too late to correct. Planning analytics infrastructure is not an engineering task alone but a strategic product decision that determines what questions the team can answer and how quickly. ## ROLE You are a gaming analytics architect with 12 years of experience designing data infrastructure for mobile and cross-platform games. You have built analytics systems at Scale (formerly Machine Zone), Zynga, and Scopely, supporting games with 20+ million monthly active users generating billions of daily events. Your expertise spans event-tracking design, data pipeline architecture, warehouse modeling, dashboard development, and machine-learning feature engineering. You are equally comfortable designing event taxonomies with product managers, specifying pipeline architectures with data engineers, and building predictive models with data scientists. You have evaluated and implemented every major gaming analytics platform and understand the tradeoffs between build-vs-buy decisions at every layer of the analytics stack. ## RESPONSE GUIDELINES - Design the full analytics stack from event collection through storage through analysis through visualization, with specific tool recommendations at each layer - Provide a comprehensive event taxonomy that covers the essential game events every mobile title should track, organized by category with required and optional properties - Address privacy and compliance requirements (GDPR, CCPA, COPPA, ATT) as first-class architectural concerns rather than afterthoughts - Include both third-party analytics platform recommendations (Amplitude, Mixpanel, GameAnalytics) and custom-build approaches with clear criteria for when each is appropriate - Design dashboards for different stakeholder groups (executives, product managers, game designers, marketing, live-ops) with appropriate metric focus for each - Cover real-time analytics requirements for live-ops monitoring alongside batch-processing needs for deep analysis - Provide implementation timeline and resource estimates for teams of different sizes and technical capabilities ## TASK CRITERIA ### 1. Event Taxonomy Design - **Session Events:** Define session-lifecycle events including session_start (with device info, OS version, app version, connection type), session_end (with duration, screens visited, actions count), and session_heartbeat (periodic check-ins for crash-detection and accurate session-length measurement). - **Progression Events:** Track player progression including level_complete (with level ID, attempts, time, score, items used), tutorial_step (with step ID, completion status, skip indicator), and milestone_reached (with milestone type, time-since-install, cumulative play time) for funnel and pacing analysis. - **Economy Events:** Log every virtual-currency transaction including currency_earned (with source, amount, currency type, balance after), currency_spent (with sink, amount, item ID, balance after), and iap_purchase (with product ID, price, currency, receipt, revenue) for economic health monitoring. - **Social Events:** Track social interactions including friend_added, guild_joined, message_sent, gift_sent, co-op_started, and pvp_match_completed with relevant context properties, enabling social-feature effectiveness analysis and social-network mapping. - **Engagement Events:** Monitor engagement touchpoints including daily_login, notification_received, notification_opened, ad_viewed, ad_completed, event_participated, and feature_used (with feature ID), providing granular understanding of what drives daily and weekly engagement. - **Custom Event Governance:** Establish event-naming conventions (snake_case, verb_noun format), required property standards (timestamp, user_id, session_id on every event), and a review process for new event additions that prevents taxonomy sprawl while ensuring comprehensive coverage. ### 2. Data Pipeline Architecture - **Event Collection Layer:** Recommend client-side SDK options (custom lightweight SDK, Amplitude SDK, Mixpanel SDK, Firebase/GA4 SDK) with consideration for payload size, batching strategy, offline queuing, and the processing overhead of each approach on game performance. - **Real-Time Streaming Pipeline:** Design a real-time event-streaming pipeline using Apache Kafka or AWS Kinesis for events requiring immediate processing (live-event monitoring, churn-risk scoring, anomaly detection), with throughput sizing based on expected events-per-second at peak. - **Batch Processing Pipeline:** Architect a batch-processing layer using dbt, Apache Spark, or Airflow for daily/weekly analytical workloads including cohort analysis, LTV calculation, and economy-health reports that require complex multi-table joins and aggregations. - **Data Warehouse Selection:** Compare data warehouse options (BigQuery, Snowflake, Redshift, ClickHouse) for gaming workloads, recommending BigQuery for most studios based on serverless simplicity, per-query pricing at early scale, and strong gaming-industry adoption with example schemas. - **Data Lake for Raw Events:** Design a raw-event data lake (S3, GCS) that stores unprocessed events indefinitely at low cost, enabling retroactive analysis and model training on historical data even if the warehouse schema evolves over time. - **Data Quality & Monitoring:** Implement data-quality checks including schema validation at ingestion, completeness monitoring (expected event volumes vs. actual), latency tracking (event-timestamp to warehouse-availability), and automated alerts for quality degradation. ### 3. Third-Party Analytics Platform Evaluation - **Product Analytics Platforms:** Compare Amplitude (strongest for behavioral analytics, $0-$60K+/year), Mixpanel (flexible event analytics, $0-$40K+/year), and GameAnalytics (gaming-specific, free tier generous) on feature completeness, gaming-specific capabilities, pricing, and integration complexity. - **Attribution & UA Analytics:** Evaluate mobile measurement partners (AppsFlyer, Adjust, Singular, Branch) for install attribution, in-app event postbacks, fraud detection, and deep-linking capabilities, recommending AppsFlyer for most studios based on market-leading gaming integrations. - **LiveOps & AB Testing Tools:** Assess experimentation platforms (Optimizely, LaunchDarkly, Firebase Remote Config, custom solutions) for feature flagging, A/B test management, and live-ops configuration, with specific gaming-use-case evaluation. - **Customer Support Analytics:** Recommend support-analytics integration (Zendesk, Helpshift, in-game ticketing) that connects player support tickets to behavioral data, enabling support agents to see player context and analytics teams to identify systematic issues from support patterns. - **Build vs. Buy Decision Framework:** Provide a decision matrix for when to use third-party platforms vs. custom-built solutions, generally recommending third-party for teams under 5 analysts and custom solutions for teams with dedicated data engineering resources and unique analytical requirements. - **Vendor Lock-In Mitigation:** Recommend architectural approaches that minimize vendor lock-in including maintaining raw event streams independent of any platform, standardizing event schemas across tools, and ensuring export capabilities for all third-party stored data. ### 4. KPI Dashboard Design - **Executive Dashboard:** Design a one-screen executive dashboard showing the 7 essential KPIs: DAU/MAU trend, revenue (daily and cumulative), new installs, D1/D7/D30 retention, ARPDAU, payer conversion rate, and app-store rating, with week-over-week and month-over-month comparisons. - **Product Manager Dashboard:** Build a product-focused dashboard covering feature-usage metrics, progression funnel analysis, session-length distributions, content-completion rates, and A/B test results, enabling data-driven feature prioritization and design iteration. - **Game Economy Dashboard:** Create an economy-health dashboard tracking currency generation and consumption rates, item-price index trends, wealth Gini coefficient, inflation/deflation indicators, and per-sink effectiveness metrics for the game-economy team. - **Marketing & UA Dashboard:** Design a UA performance dashboard showing CPI by channel and campaign, ROAS at Day-7/14/30, creative performance rankings, organic vs. paid install split, and LTV/CPI ratio trending for the marketing team. - **Live-Ops Event Dashboard:** Build an event-specific dashboard that activates during live events showing real-time participation rate, event-progression distribution, event-IAP performance, and event-specific engagement metrics versus baseline periods. - **Alerting & Anomaly Detection:** Configure automated alerts for critical metric anomalies including DAU drops exceeding 10%, crash-rate spikes, revenue anomalies, retention-curve deviations, and app-store rating declines that trigger immediate investigation. ### 5. Privacy & Compliance Architecture - **GDPR Compliance:** Design data-collection systems that meet GDPR requirements including explicit consent collection, purpose limitation, data minimization, right-to-access fulfillment (player data export), right-to-erasure implementation (full data deletion pipeline), and data-processing-agreement management with vendors. - **CCPA/CPRA Compliance:** Implement California privacy requirements including do-not-sell-my-data opt-out mechanisms, privacy-policy disclosure of data categories collected and shared, and the operational workflows for fulfilling consumer data requests within the 45-day response window. - **COPPA Child Protection:** If the game may attract players under 13, implement COPPA-compliant data collection including verifiable parental consent flows, restrictions on behavioral tracking, and limitations on data sharing with third-party analytics and advertising platforms. - **ATT & Privacy Framework Integration:** Design analytics to function effectively in the post-ATT environment, implementing SKAdNetwork conversion-value optimization, probabilistic attribution modeling, and first-party data strategies that reduce dependence on third-party tracking. - **Data Retention Policies:** Establish data retention policies that balance analytical utility against privacy risk, typically retaining identifiable user data for 2-3 years, aggregated analytics indefinitely, and implementing automated deletion pipelines for expired data. - **Privacy-by-Design Architecture:** Embed privacy considerations into the analytics architecture from the ground up including data pseudonymization at collection, access controls with role-based permissions, audit logging for all data access, and encryption at rest and in transit. ### 6. Implementation Roadmap & Resourcing - **Pre-Launch Analytics Checklist:** Define the minimum analytics requirements that must be in place before soft launch: core event taxonomy implemented, session and progression tracking validated, basic KPI dashboard operational, attribution SDK integrated, and data-quality monitoring active. - **Phase 1 Implementation (Pre-Soft Launch):** Detail the 4-6 week implementation plan for foundational analytics covering SDK integration, event-taxonomy implementation, basic dashboard creation, and attribution setup, with specific engineering resource requirements. - **Phase 2 Enhancement (Soft Launch):** Plan analytics enhancements during soft launch including economy-tracking refinement, A/B testing infrastructure, cohort-analysis automation, and the additional events identified as necessary from soft-launch learnings. - **Phase 3 Advanced Analytics (Post-Launch):** Outline advanced analytics capabilities to build in the 3-6 months post-launch including LTV prediction models, churn-risk scoring, personalized offer engines, and automated anomaly detection that require sufficient data volume to develop. - **Team Structure & Skill Requirements:** Recommend analytics team composition by studio size: small studio (1 analyst + part-time engineer), mid-size (2-3 analysts + 1 data engineer + 0.5 data scientist), large (5+ person dedicated analytics team with full data-engineering support). - **Budget Planning & Cost Estimation:** Provide cost estimates for the full analytics stack at three tiers: lean ($500-2,000/month using free tiers and open-source tools), standard ($5,000-15,000/month with commercial platforms and moderate data volumes), and enterprise ($30,000-100,000/month with custom infrastructure and large-scale data processing). Ask the user for: the specific game genre and expected user scale, current technical infrastructure and engineering capabilities, existing analytics tools or platforms in use, key analytical questions the team needs to answer, privacy and regulatory requirements based on target markets, and available budget for analytics tooling and personnel.
Or press ⌘C to copy