You are about to launch. Or you launched last week and you are refreshing App Store Connect like it owes you money.
Two things happen in those first 24 hours, and most launch advice mixes them up.
And somewhere in your head is a story about launch day. Apple is watching. The first 24 hours set something in motion. Get the push right inside that window and the algorithm pays you back.
I believed a version of that too. An earlier draft of this article said so in print.
The first is measured, and it is enormous. Across 115,000+ apps, RevenueCat finds that nearly all trial starts, roughly half of all paid conversions and most trial cancellations land on Day 0, the day someone installs. Your users decide almost immediately.
The second is what Apple does with any of it, and it is smaller and far more boring than the folklore. Apple publishes one sentence about how search results are ranked. Most of what gets written about the launch-day algorithm, including in that earlier draft, is not in it.
So the mistake is spending launch week on the algorithm instead of on the screen your users actually meet.
What follows: the Day 0 numbers by category, Apple’s sentence in full so you can judge it, six apps that put a dated trial timeline in front of people on install day, and the two claims I deleted from this page.
This is the part with numbers behind it. All of it comes from RevenueCat’s State of Subscription Apps 2026, which measures more than 115,000 apps.
Share of each event that happens on the day of install, not spread over the days after it.
Trial starts is a range because it moves by category: 89.9% in Business, 82.1% in Health & Fitness, 81.5% in Gaming, 78% in Productivity. The bar shows the middle of that range. RevenueCat, State of Subscription Apps 2026, 115,000+ apps.
In most categories, between 78% and 90% of all trial starts happen on the day of install. 89.9% in Business, 82.1% in Health & Fitness, 81.5% in Gaming, and 78% in Productivity, which is the lowest of the four the report singles out. Days 1 to 3 pick up another 5 to 12%.
After Day 3, trial starts fall below 5% everywhere. The report draws the conclusion itself: users who don’t try immediately rarely try at all.
50.6% of download-to-paid conversions happen on Day 0. That rises to 71.9% in Productivity and falls to 28.5% in Education, where 7-day trials are common and push the conversion out to the trial-end mark.
Hannah Parvaz, founder of the growth agency Aperture, puts it plainly in the report’s own commentary: if you want to make money in 2026, start with Day 0.
| Trial length | Cancellations on Day 0 |
|---|---|
| 3 days | 55.4% |
| 7 days | 39.8% |
| 14 days | 35.7% |
| 30 days | 31.1% |
Share of trial cancellations that land on Day 0, by trial length. RevenueCat, State of Subscription Apps 2026, p. 61. Read that column carefully: these are shares of the cancellations, not shares of the people who started a trial. They tell you when the decision to leave gets made, not how many people make it.
Widen the window by one day and 84% of 3-day and 64% of 7-day trial cancellations have already happened.
Put the three together and the first 24 hours are not a marketing window. They are the window in which almost everything gets decided: whether someone tries, whether someone pays, and whether someone walks.
Apple publishes one sentence about search ranking, on its own developer site:
“Search results are based on a number of factors, including text relevance (matches for your app’s title, subtitle, keywords, and primary category), as well as user behavior (downloads, ratings and reviews, and more).”
What is in it: downloads, ratings and reviews. What is not in it: revenue per download, next-day retention, delete rate, per-hour velocity relative to your category, or any time window at all. The “and more” is doing real work, and nobody outside Apple knows what it covers.
An earlier version of this article made two claims. That Apple weighs revenue per download, because it takes a cut of every transaction. And that it throttles the visibility of apps whose users delete on the first day.
I could not find a source for either that I am willing to stand behind. So both are gone, along with a reference that pointed at an analytics vendor’s homepage rather than at any actual finding.
If you have met those claims elsewhere, have a look at whether anybody cites anything. In my experience they do not.
None of that makes the launch window less important. It moves the reason. The first 24 hours matter because of what your users do in them, which is measured, rather than because of what Apple is rumoured to do with them, which is not.
The usual argument is that a $0 revenue signal tells the algorithm your app doesn’t monetise. Drop that argument. The real one is arithmetic and it is worse.
If there is nothing to buy on Day 0, you are not deferring that revenue to the week you finally ship the paywall. You are mostly losing it for everyone who has already downloaded, because by then the majority of them are gone. You can add the paywall later; you cannot re-run Day 0 for the people who already installed.
The same holds for screenshots, ASO and onboarding, every element that affects conversion is being tested on your first cohort, the group of users who all arrive at once, whether it is finished or not.
If the decision lands on the day of install, one screen carries most of it: the one that introduces the trial. Reading through the paywalls in my teardown library, what the larger apps have converged on is not a better argument for subscribing.
It is a calendar. Today, reminder day, charge day, written out on the screen the user meets in their very first session, while the app is still open in front of them.
Under “Try it for 30-days free”: “Today: Unlock every single Strava feature · In 28 days: We’ll remind you that your free trial is ending soon · In 30 days: Your annual subscription begins!” The same paywall carries the line “You won’t be charged until Jul 1. Cancel any time up to 24hrs before your trial ends.”
Both the charge date and the cancellation deadline, stated before the card. Nothing about the next month is left for the user to find out later.
Headed “Free trial guide”: “TODAY: No payment now, just full access to the Premium experience · DAY 5: Get a reminder about the end date of your free trial · DAY 7: Cancel in advance to avoid payment.” Read what Day 7 says. It tells the user how to not pay, on the screen that is asking them to start paying.
That is a Day 0 decision about how much you trust the product to earn the second week.
“Today: Get instant access · Day 12: Reminder · Day 14: Trial ends”, followed by “Your paid Plus membership starts on November 4, 2024.”
“In 14 days” is an interval the reader has to compute. A written-out date is one they can put in a calendar in the session they are already in, which is the only session most of them will give you.
“Today: Try Deepstash Pro · Day 5: Trial reminder — We’ll remind you about your trial via notification. Cancel anytime in 15 seconds. · Day 7: Trial ends.” The install step above it is already ticked off, and the button reads “I’m ready”.
Most paywalls promise that cancelling is possible. This one puts a number on how long it takes. Then it has to survive contact. A fifteen-second claim that turns out to be five screens comes back as a review.
The four steps run “Completed: You’ve successfully created your profile · Today: Unlock access to over 700 CBT sessions · Day 5: We’ll send you a reminder about your trial. Cancel in less than 1 minute. · Day 7: Your free trial ends and you will be charged.” Under the last step, in the same plain type: “Forget to cancel? No problem. Get a full refund, no questions asked.”
That line answers the objection the timeline itself just raised, on the screen that raised it, and it is paid for only by people who were never going to convert. Make them chase it through an email chain, though, and you have lost. A promise made in small type and honoured slowly costs more in support tickets and one-star reviews than the cancellations it prevented.
The pattern is not niche and it is not new. 5 Minute Journal runs “Today: Start your personal growth journey · In 6 Days · In 7 Days: Full access starts. Cancel anytime before.” Mimo runs the same three beats over a fortnight and adds “No charges yet. Cancel anytime on the App Store.” Headspace, Runna, Buddy, Brilliant and Spark Mail all ship a version of it.
It is the rare pattern that raises trial starts by being more honest rather than less.
The onboarding progress bar is still running along the top of the trial screen, roughly five sixths full. Underneath it, the timeline:
“Today: Get full access to all our features · In 12 days: Get a reminder about when your free trial ends · In 14 days: Your subscription starts, unless you cancel anytime before.”
Keeping that bar running removes the seam. The price stops being a new screen arriving and becomes the next step of something you already started.
Read the dates against Kitchen Stories, though, and notice what the same screen costs. “In 12 days” is arithmetic the reader has to do; November 4 is something they can put in a calendar while the app is still open.
One caveat before you copy the layout.
Alice Muir, a subscription consultant previously at Phiture, says every one of her clients has tried this paywall, and some get frustrated when it does nothing for them.
Her reading is that the timeline sells a relationship. So it works for habit products and does very little for a utility, where somebody wants one job done now and a promised push notification on day two is no use to them at all.
Ask what your app is promising before you borrow the screen that promises a routine.
A version of this article claimed Apple reads an annual subscription as a stronger signal than a monthly one, and cited RevenueCat for it. RevenueCat does not say that, and neither does Apple. It is cut.
There is still a case for leading with the annual plan, and it is about your own economics rather than a message to the store. Yearly plans hold on to a median 20–40% of subscribers at twelve months. Monthly plans hold 6–14%, and weekly plans 1–2%, which the report describes as essentially total churn, meaning nearly everybody is gone.
Travel leads on yearly retention at 39%; Productivity lags at 23%. That gap, not an algorithm, is why an annual subscriber is worth more.
Know what you are pushing against, though. Annual plans fell from 41.4% of all subscription durations last year to 33.6% this year, while monthly and weekly rose.
Daphne Tideman, a growth advisor, reads the same report and calls it commitment reluctance. Her argument is that plan duration, trial offers and pricing all have to line up with how long your app takes to prove itself, rather than with whatever is in fashion. And that improving activation beats pushing aggressive pricing.
Don’t launch without it. Test the purchase flow on TestFlight with sandbox accounts. Make sure the paywall appears at the right moment in onboarding and that every plan you offer actually completes a purchase.
This is the item the Day 0 data promotes to first place. If somebody has to come back tomorrow to understand what your app is for, the numbers say they mostly will not.
So onboarding has to deliver enough value that the paywall lands while the user is still in the session they installed in. And if you are running a trial, the timeline has to be on the screen where it lands.
Your first screenshots are competing against established apps in the same search results, and they are the only part of your app most people will ever see. Analyse what the top four in your category are showing before you design yours, then differ from them deliberately rather than by accident.
Title, subtitle, keyword field and primary category are four of the five things Apple actually names in that sentence. So this is the part of launch prep with the most explicit backing there is. You can refine it later, but launching with placeholder metadata wastes early search impressions.
Ratings and reviews are the other thing Apple names, so early reviews are worth having. Apple publishes no figure for how much they move a ranking, and neither will I. What Apple does state, in the App Review Guidelines, is the limit:
“If we find that you have attempted to manipulate reviews, inflate your chart rankings with paid, incentivized, filtered, or fake feedback, or engage with third-party services to do so on your behalf, we will take steps to preserve the integrity of the App Store, which may include expelling you from the Apple Developer Program.”
So: ask people who have genuinely used the app, use the standard in-app review prompt, and offer nothing in exchange. A rounded-up list of friends leaving five stars on an app they never opened is the practice that paragraph is written about.
If your app is already live without a paywall or with a weak onboarding, you have a few options.
Add the paywall, fix the onboarding, optimize the screenshots, ship. Every new install still arrives with its own Day 0, and that is the cohort the fix is for. The users you already have are the ones you have lost the window on.
If you are rebranding anyway, with a new name, icon and entity, a new app record does start with no ratings and no history.
But that is a fact about the record, not a documented algorithmic reset. I have no source for the fresh boost people describe, so treat it as a side effect rather than a strategy.
It only makes sense if the rebrand was happening regardless. Don’t create a duplicate app to game the store; that is its own guideline violation.
Downloads are one of the factors Apple names, and a download from an ad counts as a download. Paid acquisition, social, partnerships and press all still work. They just cost money instead of being free.
AI Decorator launched into a category already dominated by large players and reached 32% trial-to-paid conversion from release (June to September 2025). Trial-to-paid, which is the share of free trials that turn into paying subscriptions.
None of that was decided on launch day. Before a single screen was designed, we reviewed more than 10 top-performing AI home design apps and used Sensor Tower to estimate their revenue, which ran from $400K to over $2M a month.
Then we tore down the top four in detail: the AI tooling, the business model, and what sat behind the paywall. Features with weak AI results got cut rather than shipped.
The app went live with its paywall, its onboarding and its store presence already finished. That is what the first 24 hours actually measure.
Where the numbers come from. Day 0 trial starts, p. 56. Trial cancellations by day, p. 61. Hannah Parvaz’s commentary, p. 66. Time-to-paid by category, p. 70. Year-one retention by plan duration, p. 106. Plan-duration mix and Daphne Tideman’s commentary, p. 55. All from RevenueCat, State of Subscription Apps 2026. Apple’s wording is from App Store search and the App Review Guidelines.
The Strava, Headway, Kitchen Stories, Deepstash, 5 Minute Journal and Mimo lines are transcribed from their paywalls in my own teardown library. Wording changes, so read them as observations rather than as current fact. Alice Muir’s comment is from the Botsi Price Power Playbook, p. 23.
No, but not for the reason usually given. Around half of all download-to-paid conversions happen on Day 0, and nearly all trial starts do (RevenueCat, State of Subscription Apps 2026, across 115,000+ apps). If there is nothing to buy on the day someone installs, you are not deferring that revenue, you are mostly losing it for that cohort. You can ship the paywall later. You cannot re-run Day 0 for the people who already downloaded.
One sentence, about search: results are based on “a number of factors, including text relevance (matches for your app’s title, subtitle, keywords, and primary category), as well as user behavior (downloads, ratings and reviews, and more).” Downloads, ratings and reviews are named. Revenue per download, next-day retention and a one-time launch-day boost are not, and I could find no Apple source for them.
Treat it as a retention decision about your own economics, not a signal you send to Apple. Yearly plans retain a median 20–40% of subscribers at twelve months against 6–14% for monthly and 1–2% for weekly (RevenueCat, 2026). Against that: annual’s share of all plan durations fell from 41.4% to 33.6% year over year, so highlighting it runs against what users are currently choosing.
The pattern the larger apps have converged on is a dated timeline rather than an argument. Strava: “Today: Unlock every single Strava feature · In 28 days: We’ll remind you that your free trial is ending soon · In 30 days: Your annual subscription begins!” Headway heads the same three steps “Free trial guide” and ends on “Cancel in advance to avoid payment”. Kitchen Stories writes out the real date the paid membership starts. Deepstash quantifies the exit: “Cancel anytime in 15 seconds”. One caveat from Alice Muir, subscription consultant, previously at Phiture: the timeline sells a relationship, so it suits habit products and does little for a utility.
Conversion design for subscription apps.
Screenshots, onboarding and paywalls, connected around one goal: getting more users to pay.