Churn Rate

Churn Rate is the share of users who didn't come back during a given window. It's the mirror image of retention — computed as (1 − Retention) — and often easier to communicate intuitively to non-data stakeholders. The number is the same; the framing is what changes the conversation.

Key takeaways

Formula: 1 − Retention (same window)
Reading: D1 Churn 65% means 65% did not return the day after first visit
Use: "Churn" framing often lands better than "retention" in exec conversations
See also: The inverse of [[day1-retention]] / [[day7-day30-retention]]

Cast

Mallow

SENIOR CONSULTANT · 13Y

ZehnStudio26's senior consultant. Skilled at breaking down complex topics.

Marsh

ROBLOX BEGINNER · READER VOICE

A brand-side marketer. Asks "what is that?" so the reader doesn't have to.

Retention from the other side

Marsh

How is Churn Rate different from Retention?

Mallow

It's the same number, read from the opposite direction. Retention 35% = Churn Rate 65%. Not a new measurement — just a different way to phrase it.

Marsh

So why bother having both?

Mallow

Different listeners hear them differently. Saying "D1 Retention is 35%" to an exec carries less weight than "65% drop off after the first day." Conversely, when you ship a win, "retention is up 8 points" feels positive. Same number, different temperature.

Marsh

Is monthly churn a number people care about, or just daily?

Mallow

Both. Monthly churn is the one execs anchor on because it compounds visibly. Daily churn drives the design discussion — where in the first session people leave.

Marsh

So the levers are the same?

Mallow

Same set. But Churn framing tends to pull the analysis toward why people left — exit surveys, looking at the last session before drop-off, stage-by-stage churn (tutorial drop, purchase-screen drop, etc.). Useful for diagnostic deep dives.

Formula & benchmarks

Churn Rate = 1 − Retention
Example: D1 Retention 35% ⇔ D1 Churn 65%

D1 Churn 80%+: Severe drop-off. First-session design is the bottleneck
D1 Churn 60–75%: Roblox median range — typical, not great
D1 Churn 40–55%: Strong (matches D1 Retention 45–60%)
Monthly Churn 30%: One-third of the audience leaves per month. Over six months, fewer than 12% of a cohort remains if churn stays flat

Worked example

Marsh

Our monthly churn rate is 30%. That doesn't sound terrible.

Mallow

Compounded over six months, only 12% of your starting cohort is left. 70% × 70% × 70% × 70% × 70% × 70% = 0.117. That's the trap with churn.

Marsh

12% from 100%. So if 10,000 people joined in January, by July only 1,200 of them still play?

Mallow

Right. To grow MAU, you need new arrivals to outpace that decay. Either drop churn or accelerate acquisition. Cutting monthly churn from 30% to 20% means 6-month survival jumps from 12% to 26%. More than double.

Marsh

So small churn changes have huge downstream effects.

Mallow

Exactly. That's why one-percentage-point retention wins are worth shipping.

Reading it in Roblox

Roblox Creator Hub doesn’t display Churn Rate directly. You derive it from the Retention chart in the Analytics tab: 1 minus the displayed D1/D7/D30 number gives you the corresponding churn. For monthly churn at the cohort level, use Cohort Analysis: take a month’s first-visit cohort and check what share is still active 30 days out. External tools like RoMonitor will compute compound monthly churn over a window for you.

Common misconceptions

"Low churn means everyone is happy" — not necessarily: If most of your MAU only visits once a month anyway, "30-day churn" looks great while the experience is actually dead. Pair with Stickiness.
Daily and monthly churn are not interchangeable: Daily churn compounds much faster. A 5% daily churn is a 78% monthly churn (1 − 0.95³⁰). Always state the window.
Churn isn't a single number — it shifts by cohort: Paid traffic cohorts often churn faster than organic. Average churn hides that. Slice by source.

Day 1 Retention: D1 Churn = 1 − D1 Retention
Day 7 / Day 30 Retention: Each has a corresponding Churn reading
Drop-off funnel: Tutorial-drop, purchase-screen-drop, etc. — stage-level churn analysis
Cohort Analysis: Churn compared across signup months is where you spot a regression
LTV: High churn collapses LTV. The two move together