Disabling Churn Analysis (For Now)
Churn analysis was an experiment that proved too resource-heavy, so it's getting disabled to scale daily processing.
What's New? Latest Updates, Features & Enhancements.
Churn analysis was an experiment that proved too resource-heavy, so it's getting disabled to scale daily processing.
Churn analysis now covers four playtime cohorts, blending refund risk with long-tail churn so you can see when and why players peel off.
Game detail pages now ship with a dedicated Stats and Refunds Analysis tab that breaks down refund intent, churn drivers, and modeled revenue at risk.
Game detail pages now include a Statistical Analysis block built from the full Steam review feed: playtime × sentiment, velocity, and recent monthly trends.
Use the new blend view to pick several games you already love and instantly surface Steam titles that share the same vibe.
Discover what type of game you're looking at and what features players want most. VaporLens now automatically extracts game categories and identifies the most requested features from player reviews.
Dive deeper into your favorite games. VaporLens now analyzes DLC reviews, giving you the same powerful insights for expansions as you get for base games.
Prefer to listen to the game review instead of reading it? VaporLens now lets you listen to the game review as a podcast.
Ever wondered *why* players feel a certain way? VaporLens now goes beyond just showing emotions to explaining the reasons behind them, giving you a richer understanding of player feedback.
Ever wondered how players truly feel about a game? VaporLens now surfaces the most common emotions expressed in reviews, giving you a deeper, more nuanced understanding of player sentiment.
Track how game reviews change over time with historical Snapshots. VaporLens now lets you request fresh review data and see a game's reception journey, not just a single view.
Describe games in your own words – no titles needed. VaporLens's AI hunts matches by theme, mood, and mechanics, then explains why they fit – even for foggy "that one with the zombie clowns?" memories.
Adding Game Chat to game pages using natural language processing of review data to provide instant answers to specific questions about gameplay, performance, and player feedback.
Adding Similar Games using embeddings from review data to recommend titles based on specific gameplay and narrative traits, rather than generic tags.
Minimizing duplicate summaries and contradictions in Steam review analysis. Updating importance ranking to span categories, prioritizing key player feedback in a unified view.
Adding new Smart Sampling feature that prioritizes high-impact Steam reviews while filtering out noise. It selects up to 2000 reviews based on recency, detail, playtime, and community upvotes for a focused view.