What do D1, D7 and D30 retention mean on Roblox?
Roblox defines D1 retention as the share of new users who played again after their first session ended. What each window tells a brand, and which one to judge a world on.
Retention is the number that separates a world people visited from a world people use. Roblox reports it on its own definitions, and reading them precisely saves a lot of argument later.
The definitions, as the platform states them
D1 retention: the percentage of new users who played again after their first session ended.
D7 and D30: the same question over seven and thirty days.
New users: the daily active users who are first-time users.
DAU: unique players who joined at least once in the given day.
Note what D1 is counted against. The denominator is new users, so a traffic spike from a campaign lands in the denominator immediately and in the numerator only if those people come back.
Why a launch push makes retention look worse
Buy a burst of traffic and D1 falls, because you added a large group of first-timers with no particular reason to return. The world did not get worse. The mix changed.
Read retention alongside where the traffic came from, and read it again a week after the push ends.
What each window is good for
D1 tells you whether the first session worked: whether people understood the loop and had a reason to come back tomorrow. This is a design signal, and it is the one to fix first.
D7 tells you whether there is enough content and enough reward depth to survive a week.
D30 tells you whether the world has a life beyond the campaign. For an always-on presence this is the number that matters; for a dated promotion it will be low and that is expected.
The brand-side mistake
Treating retention as a vanity number because visits look better in a report. Visits measure distribution. Retention measures the product. A world with large visits and collapsing retention is a media buy pointed at an empty room.
What to ask for
Retention split by acquisition source, so an organic cohort can be compared with a paid one. The two behave very differently, and averaging them hides the only useful comparison.
Retention before, during and after any push, since the campaign period on its own is unreadable.
Setting a target honestly
There is no published cross-industry benchmark for branded worlds that we would put in a proposal. What works is a baseline: measure the pilot cohort, then set the full build’s target against your own number. Anyone quoting you an industry retention figure for branded Roblox worlds should be asked where it was published.
Roblox game development for brands that need it to work
Hiring a freelance Roblox developer gets you a place. Getting a world that people return to is a production problem — gameplay, economy, live-ops and analytics designed together, before the first block is placed.