Related engagement
Analytics & KPI Framework
A decision-grade metric architecture spanning definitions, ownership, dependencies, review cadence, and action thresholds.
Explore the engagement →Topic 03 · Decision systems
Gaming expertise. Impact that travels.
Analytics & KPI.
More data does not automatically create better decisions. A useful analytics system makes definitions explicit, preserves cohort context, exposes dependencies, and tells a team when the evidence is mature enough to act. Gaming is the proving ground here; the same discipline applies to app, SaaS, MMP, and analytics-platform teams.
Diagnostic frame
Start with the questions behind the metric.
- Does the KPI model reflect real dependencies, or just a tidy hierarchy?
- Which dashboard signals are causal, lagging, or simply immature?
- Can two teams reproduce the same answer from the same definition?
Selected writing
Read the field notes
- Sep 2026 D—03 · Analytics 3 min
AppsFlyer vs Adjust vs Singular: Choosing an MMP for a Mobile Game
Every major MMP attributes installs and supports SKAdNetwork. The differences that matter for games are in cost and ROI aggregation, raw data access, pricing model, fraud tools and ownership.
- Sep 2026 D—03 · Analytics 3 min
Bookings vs Revenue in Games, Explained
Bookings tell you what players spent this period. Revenue tells you what accounting recognises. Confusing the two is how teams misread momentum in both directions.
- Sep 2026 D—03 · Analytics 3 min
The Game Growth Tool Stack: MMPs, Analytics, Market Intelligence, ASO and CRM
A map of the tools most game growth teams use, what each category is for, and how to avoid buying five tools that all answer the same question.
- Sep 2026 D—03 · Analytics 3 min
How to Calculate LTV for a Mobile Game
LTV is retention multiplied by revenue per active day, summed over a horizon. The arithmetic is easy; the judgment is in the tail of the curve you have not observed yet.
- Sep 2026 D—03 · Analytics 3 min
Using LLMs for Game Analytics: A Practical Workflow (and Where It Fails)
Large language models can write SQL, summarise cohorts and draft analyses fast. They are only as reliable as the definitions you give them — so the metrics dictionary becomes the most important prompt you own.
- Sep 2026 D—03 · Analytics 2 min
The Averages Trap: Why Cohort Blending Hides Burning Markets
Blended D30 can stay flat while the market receiving your next UA dollar deteriorates. An average is useful for reporting; it is not automatically the right number for allocation.
- Sep 2026 D—03 · Analytics 2 min
A Dashboard Should Compress a Decision, Not Display the Database
Most dashboards answer a question the operator did not ask: what data do we have? The useful question is what changed, who decides, and what they can do now.
- Aug 2026 D—03 · Analytics 2 min
Recurrent Revenue Percentage Is Composition, Not Health
Take-Two reported recurrent consumer spending at 84% of net revenue in Q1 FY2027. That number describes revenue mix — it cannot tell you whether the quarter was healthy.
- Aug 2026 D—03 · Analytics 2 min
Bookings and Revenue Run on Different Clocks
Roblox reported $4.89B of 2025 revenue and $6.79B of bookings. The gap is not an error. It is a clock.
- Aug 2026 D—03 · Analytics 2 min
A Metrics Dictionary Is a Product, Not a Document
“Retention is up” can be true and still useless if the team hearing it and the team writing it mean different things.
- Jul 2026 D—03 · Analytics 2 min
A Recommendation Should Come With an Expiry Condition
Teams document the recommendation and the numbers behind it. The assumption carrying it usually stays in the meeting — so nothing reopens the decision when the evidence changes.
- Jun 2026 D—03 · Analytics 7 min
More Data, Less Signal
More channels don't just add inventory — they dilute signal until you're reconciling data instead of making decisions.
- May 2026 D—03 · Analytics 4 min
When Market Intelligence Becomes Infrastructure
As market intelligence consolidates, the advantage stops coming from having the tool. It comes from the internal intelligence you build around it.
- Apr 2026 D—04 · Revenue 8 min
Your KPI tree is probably a graph
Why North Star frameworks quietly produce the wrong roadmap, and the cheap fix.
- Oct 2025 D—02 · Apps 9 min
When to A/B test, and when not to
Most teams under-experiment on pricing and over-experiment on copy. Here's the heuristic I use.
Need an outside operator’s view of the system behind the number?
Start the conversation →