Before you begin
For better results, tell the AI assistant:
What you want to investigate
Which games, genre to include
The relevant date range
The metrics or evidence you need
How you want the answer presented
Note: Responses may vary by AI model and application. Some applications can display screenshots directly, while others return links. Coverage also depends on the selected game, market and period.
1. Explore and compare
Quickly find relevant games, features and LiveOps events. You can request implementation examples, analyst commentary, screenshots and direct links to the corresponding GameRefinery pages.
Prompt example: examine ad placements
Use GameRefinery to find and display screenshots of ad placements and corresponding analyst comments (if available) from Bus Traffic Fever! Render the screenshots so I can see them directly in the agent.
Prompt example: analyse anniversary strategies
Use GameRefinery to analyse how the four highest-grossing shooter games on iOS in the US, based on the latest available ranking, approached their most recent anniversary updates.
Compare:
Update timing and duration
Application update size
New content, features and LiveOps events
Anniversary rewards and monetisation
IP collaborations or major campaigns
Available revenue, download and active-user changes
Summarise the common approaches, notable differences and strongest measurable performance changes. Include links to the supporting GameRefinery event or update pages and distinguish observed correlation from causation.
2. Analyse events changes and their potential impact
Combine event or update data with available KPIs to investigate which LiveOps implementations appear to have performed well.
Important: Performance changes should normally be treated as correlations rather than proof that an event caused the observed result. Other events, updates, marketing activity and external factors may affect performance during the same period.
Prompt example: connect event changes with game performance
Use GameRefinery to analyse how Royal Match’s LiveOps events changed during the past 12 months.
Summarise the most significant changes to:
Event mechanics
Themes
Rewards
Event cadence
Compare the timing of these changes with available revenue, download and active-user data. Highlight the strongest relationships, distinguish correlation from confirmed causation, and include links to the relevant event descriptions.
Prompt example: measure the impact of IP collaborations
Use GameRefinery to identify the most impactful IP collaboration events in the puzzle genre during 2026 in the US market.
For this analysis, define revenue impact as:
Revenue generated during the collaboration minus revenue generated during the immediately preceding period of equal length.
For each collaboration, provide:
Game and event name
Collaboration IP
Event start and end dates
Event duration
Revenue during the event
Revenue during the comparison period
Absolute revenue difference
Percentage revenue difference
Link to the full event description
Sort the results by absolute revenue difference in descending order. Explain the methodology, identify overlapping events or updates that could affect the comparison, and clearly state when required data is unavailable.
3. Identify market trends
Examine events across multiple games to identify common durations, mechanics, themes and operating patterns.
Prompt example: compare collectible-album durations
Use GameRefinery to analyse seasonal collectible-album events in the puzzle genre during the trailing 12 months.
List the most common exact durations and group them into practical duration bands. For each duration, provide:
Number of observed album runs
Percentage of all observed album runs
Representative game examples
GameRefinery event terminology
Using GameRefinery’s event terminology can produce more precise results:
Live-event instance: One occurrence of a named event
Blueprint: A recurring event template within one game
Blueprint family: A shared event template found across multiple games (not available at the moment)
Event type: A general mechanical classification
Event category: The broadest classification layer