Growth & updates
Roblox Home PTR and early exits: find the weak stage
A fall in Home impressions does not identify the change to make. Inspect the chain in order: eligibility and delivery, play after impression, early exit, and later behavior of the organic RFY cohort.
Focused deep dive
Read the chain, not one number
Current Roblox documentation separates retrieval from ranking. Its core ranking-signal set includes PTR, first-play bounce, play days and playtime across several windows, co-play, qualified sessions, and monetization. Roblox does not publish exact signal weights. These signals use players acquired organically from Recommended for You. Sponsored CTR is a different measure.
Save the metric name and live tooltip because older qPTR material may describe a previous model. If impressions fall, align the change with PTR and bounce, but retain alternatives: platform demand, competition, and audience composition can also move distribution.
| Observation | Check | Next action |
|---|---|---|
| Stable impressions, falling PTR | Promise accuracy, legibility, segment | Test one thumbnail concept |
| Stable PTR, rising bounce | Loading, spawn, promised action | Repair the demonstrated entry blocker |
| Stable entry, weaker return | Play days, playtime, progression | Test the reason to continue |
| All averages move together | Source, country, platform, maturity | Decompose audience mix |
Worked hypothetical example
Suppose Home impressions double, PTR moves from 4% to 3%, and plays rise from 400 to 600. First-play bounce increases only among mobile players in a newly reached country. This does not establish a failed thumbnail: expansion changed the mix while absolute plays rose. Check loading and the opening screen on representative devices for that segment, then compare performance inside a matching cohort.
If source × country × platform is not available as a joint breakdown, record the limit. Separate charts do not prove that the same players create the observed problem.
- Record the window, definition, and source of every metric.
- Compare impressions, PTR, and bounce without ad metrics.
- Review mature behavior windows and supported breakdowns.
- Choose one repair or test and preserve its date as a milestone.
Primary sources
Put this into practice in MARPLA
Separate organic Home from other sources, inspect countries and supported joint breakdowns, then relate audience changes to Home set history and record a hypothesis in a note.
- Investigate Home and new audiencesCheck source, countries and cohort maturity to separate audience-mix changes from a worse game experience.Step-by-step guide →
- Compare audience composition and qualityCompare a complete period with its predecessor and separate audience mix change from within-segment metric change.Step-by-step guide →
- Compare Home creatives and setsDistinguish set changes from statistics days.Step-by-step guide →
Where this deep dive applies
- Roblox Home recommendations: get discovered and sustain growth
- Roblox thumbnails: screenshots, staged scenes and practical testing
- Roblox advertising: when to launch, how to target, and when to stop
- How to improve a Roblox game's first session and retention
- A Roblox game launch and update plan: soft launch through review
Continue exploring
More on this topic →
Roblox Home recommendations: get discovered and sustain growth
How Roblox tests and expands Home recommendations, what signals matter, and how to diagnose country mix and falling metrics with MARPLA.
Read article
Roblox thumbnails: screenshots, staged scenes and practical testing
Choose Roblox thumbnails, structure the game-page gallery and test Home images: screenshots versus renders, PTR/qPTR, creative briefs and a practical workflow.
Read article
How to improve a Roblox game's first session and retention
Move from observed playtests and an onboarding funnel to mature D1/D7 cohorts and testable changes.
Read article
Discussion0
Loading comments…