Roblox Visits vs players: counters that answer different questions

Four measures, four denominators

CCU answers how many users were simultaneously playing at an observation time. A Visits increment answers how much the cumulative visit counter increased between two snapshots; it does not reveal unique people because one user can generate more than one visit. DAU is available to an authorized owner and counts unique users who joined at least once during the day. A session is a period of play, so one DAU can have multiple sessions.

The measures can move in different directions without contradiction. A game can have lower evening CCU than yesterday while still adding more Visits across the full day. A promotion can add many brief visits and raise DAU while average session time declines. Retention needs a cohort of new users and a later return window; neither CCU nor lifetime Visits supplies that denominator.

Worked example: 12,000 new Visits is not 12,000 players

Suppose public Visits rises from 4,000,000 to 4,012,000 in 24 hours. The game’s observed CCU ranges from 180 to 420. The only public statement is that the counter increased by 12,000 during that interval. If each user played once, 12,000 users is possible; if users rejoined, the unique count is lower. Missing snapshots or counter-update timing can also shift the apparent interval.

The owner dashboard reports 8,000 DAU and 10,500 sessions for that same aligned day. Now the team can say 8,000 unique daily users generated 10,500 sessions in the owner dataset. It still should not force the public 12,000 counter increment to equal either number: definitions, time zones, processing, and counted join behavior need reconciliation first.

Choose the measure from the decision

Write the noun beside every number: concurrent users, visit-counter increment, unique daily users, or sessions.

1. Align the period and time zone before comparing sources.
2. Use CCU for concurrency shape and scale.
3. Use Visits increments as a public activity counter, not unique reach.
4. Use owner DAU and sessions for audience and frequency questions.
5. Use retention cohorts only for return behavior.

Common denominator errors

Do not divide revenue by public Visits and call the result ARPDAU. Do not estimate retention from CCU divided by Visits. Avoid summing CCU points to get unique users; the same people can appear in many observations. A raw lifetime Visits total also gives older experiences a structural advantage in rankings.

A derived session estimate can use the area under a sampled CCU curve divided by a Visits increment, but it estimates minutes per added visit under assumptions. It does not recover unique players or the official owner KPI. Keep its confidence and coverage visible.

Decision record

Write the metric noun and denominator in the decision note. If an owner export later supplies DAU or sessions, add it as a separate measure and reconcile the time zone before comparison. Never rename a public Visits increment as users merely because a stakeholder wants an audience number.

Primary sources

• Roblox Analytics
• Roblox Analytics dashboard
• Roblox games API