Most published renewal numbers are medians across thousands of apps. This post is the other end: one small app, every period counted, with the sample size next to each rate. The numbers come from FolioKit, the analytics tool I build and run my apps on. I'm not publishing prices or revenue here, only rates and counts.
What was measured
- One app. LuminaClean only. My other three apps have too few subscription periods to rate.
- A period is one paid week or one paid month. It counts once it ended at least 6 days ago, so a late renewal has time to arrive. The data starts when the RevenueCat webhook went live on July 10, 2026, and runs to September 23.
- Renewed means the next paid period arrived. First renewal is a subscriber's first paid period renewing; later renewals are every period after that.
- Cancelled means auto-renew was turned off. The
subscriber keeps access until the period ends; RevenueCat reports it
as a
CANCELLATIONevent, which doesn't mean what it sounds like. - My own test devices and pirated copies are excluded.
Renewal rates by plan
| Plan | All periods | First renewal | Later renewals |
|---|---|---|---|
| Monthly | 75.0% of 40 | 1 observed, not rated | 76.9% of 39 |
| Weekly | 47.8% of 23 | 33.3% (4 of 12) | 63.6% (7 of 11) |
The monthly plan has only one observed first renewal because most monthly subscribers started before tracking did. Its 75.0% is almost entirely later renewals.
That makes the headline comparison unfair to the weekly plan. The weekly figure mixes first renewals, where most churn happens, with later ones; the monthly figure is nearly all later renewals. The like-for-like column is later renewals: 63.6% of 11 for weekly against 76.9% of 39 for monthly.
The two rates also run on different clocks. A weekly subscriber faces about four renewals in the time a monthly subscriber faces one, so a per-period rate that looks close to monthly still loses more people per month.
23 weekly periods is not much, so treat this as a direction. With 12 first weeks, one subscriber more or less moves the first-renewal rate by several points.
Next to RevenueCat's 2026 benchmarks
RevenueCat publishes renewal medians from its State of Subscription Apps 2026 report (115,000+ apps), broken down by category in average subscription renewal rates by app category (updated April 24, 2026). LuminaClean is listed under Productivity, with Photo & Video as its secondary category. The table shows RevenueCat's data for those two categories, with my one app below for reference:
| First renewal | Weekly: median (25th–75th percentile) | Monthly: median (25th–75th percentile) |
|---|---|---|
| RevenueCat 2026, Productivity | 53% (45–65%) | 54% (40–66%) |
| RevenueCat 2026, Photo & Video | 45% (31–53%) | 48% (35–59%) |
| LuminaClean (my data) | 33.3% (4 of 12) | 1 observed, not rated |
Across RevenueCat's 11 categories, the median first weekly renewal is below the monthly one in all but Education. My weekly first renewal is below the Productivity lower quartile (45%) and close to the Photo & Video one (31%). With 12 periods behind it, that's a reason to watch the first week closely, not proof that the app is below the market. RevenueCat's exact definitions of a renewal rate may also differ from mine, so compare shapes rather than decimals. For the monthly plan there's no like-for-like comparison yet: RevenueCat's category tables cover the first renewal, and I have one.
Three findings from RevenueCat's 2026 data are useful next to mine:
- The first weekly renewal is the hard one. In RevenueCat's data, weekly renewal rates jump by roughly 25 to 30 percentage points between the first and second renewal, and by the third renewal the medians sit between 74% and 91% across categories. My weekly plan went from 33.3% (of 12) at the first renewal to 63.6% (of 11) after it.
- People cancel early. RevenueCat's 2026 summary reports that 55.4% of 3-day trial cancellations happen on day 0. Those are trials, not paid weeks, but it's the same habit my weekly subscribers show below.
- Billing failures are a minority of App Store churn. The same summary puts billing errors at 14% of App Store cancellations, against 31% on Google Play.
When people cancel
Timing of each cancellation, as a share of that plan's cancellations:
| Auto-renew turned off | Weekly (11 cancellations) | Monthly (10 cancellations) |
|---|---|---|
| Within 24 hours of paying | 63.6% | 30% |
| In between | 18.2% | 70% |
| In the last 3 days before renewal | 18.2% | 0% |
Most weekly subscribers who cancel decide at the moment they pay. They look like people who wanted one week of the app. Monthly subscribers decide during the month, and none of the 10 monthly cancellations came in the last three days before the renewal.
That timing has a useful side effect for the weekly plan. Since most weekly cancellations land within a day, auto-renew status a day after purchase already shows most of what the first renewal will look like. A change to onboarding or the paywall gets an early read the next day instead of a week later, as long as you remember it's an early read on a handful of people.
The number to plan with: renewal chance if not cancelled yet
A flat renewal rate answers "what share of periods renewed". When you
plan the rest of a month, the question is different: of the
subscriptions that are live today and still set to renew, how many
will? You already know which ones have auto-renew off, from
RevenueCat's cancellation events or from Apple's
autoRenewStatus.
Applying the flat 47.8% weekly rate to subscriptions that are still on
counts the day-one cancellations twice.
So I measure the renewal chance for subscriptions that haven't been cancelled yet, by the days left until their renewal (LuminaClean, as of September 23, 2026):
| Days until renewal | Monthly, not cancelled yet | Weekly, not cancelled yet |
|---|---|---|
| Within 3 days | 99% (30 periods) | 81% (14 periods) |
| 4 to 10 days | 91% (33 periods) | 72% (10 periods) |
| 11 days or more | 85% (34 periods) | does not apply |
The chance rises as the renewal gets closer, because there's less time left to cancel. A monthly subscription that is still on three days before renewal almost always renews, which matches the 0 of 10 monthly cancellations in the last three days. The weekly plan stays lower even close to renewal, since 18.2% of its cancellations come in those last three days.
These are the numbers my month forecast runs on. When I backtested it, cautious default renewal chances under-predicted renewals by 15–20% until the app had its own history, which is the argument for measuring your own instead of borrowing a benchmark.
Failed payments: none so far
0 of 63 periods failed on payment. With 63 periods that doesn't mean the rate is zero; RevenueCat's 14% share of App Store cancellations from billing errors is the better prior once the numbers grow. It does mean that no subscriber in this data was lost to a failed payment.
What this means for choosing and pricing plans
Only as far as this data supports:
- Judge a weekly plan by its first renewal. In this app 4 of 12 first weeks renewed and 7 of 11 later weeks did, and most weekly cancellations came within a day of paying. That first day is where to look when a weekly plan underperforms.
- Price the weekly plan on paid weeks per subscriber, not on the monthly price divided by four. If r1 is your first renewal rate and r2 your later renewal rate, a new subscriber pays for about 1 + r1 / (1 − r2) periods, assuming later renewals hold steady. RevenueCat's data says they tend to rise, so this errs low. Do the same for the monthly plan and compare price times periods. With 12 and 11 periods behind my weekly rates, I wouldn't feed them into a price decision yet.
- Forecast with conditional chances. Use the renewal chance for subscriptions not yet cancelled, by plan and days to renewal, and let weekly subscriptions renew several times before the month ends.
What this data can't tell you: which plan earns more per buyer, whether offering a weekly plan pulls people away from longer ones, or anything about yearly plans. For subscription retention by start month from Apple's side, App Store Connect now has cohorts; see App Store Connect analytics in 2026.
FAQ
Do weekly subscriptions renew less often than monthly ones?
In my app, yes: 47.8% of 23 weekly periods against 75.0% of 40 monthly periods. Most of the gap is the first weekly renewal (4 of 12); later weekly renewals ran at 63.6% of 11.
What is a good first renewal rate for a weekly subscription?
RevenueCat's 2026 medians range from 35% to 58% by category, with Productivity at 53% and Photo & Video at 45%. Rates jump by roughly 25 to 30 points at the second renewal.
When do weekly subscribers cancel?
Mostly right away: in my app 63.6% of 11 weekly cancellations came within 24 hours of paying, against 30% of 10 for monthly.
How should I forecast subscription renewals for the month?
Give each live subscription that still has auto-renew on a renewal chance by plan and days to renewal. In my app that was 99% for a monthly renewal within 3 days and 81% for a weekly one.
How many App Store subscription payments fail?
RevenueCat's 2026 data puts billing errors at 14% of App Store cancellations. My app had 0 of 63 periods fail, too small a sample to plan on.
Sources
- RevenueCat, Average subscription renewal rates by app category (2026 benchmarks), last updated April 24, 2026 (State of Subscription Apps 2026 data)
- RevenueCat, The State of Subscription Apps in 10 minutes: lessons, trends, and benchmarks for 2026, March 19, 2026, last updated April 22, 2026
- RevenueCat, State of Subscription Apps 2026
- Apple, autoRenewStatus (App Store Server API)
- My own data: LuminaClean subscription periods from RevenueCat webhooks, July 10 to September 23, 2026, test devices and pirated copies excluded