Sunil GillA soft launch is a release to a deliberately small audience, usually one or two countries, before the app goes out everywhere. You get one launch. A soft launch is how you avoid spending it on a version that crashes, confuses people or cannot earn back what its users cost.
Done properly it is three tests run in sequence, each with a pass mark you set in advance. This guide covers those tests, what the App Store and Google Play let you restrict, how to choose markets, and real benchmarks to set your targets against.
Soft launch vs hard launch
| Soft launch | Hard launch | |
|---|---|---|
| Audience | A few test markets or a capped group | Every market you plan to serve |
| Marketing | Small, measured spend to bring in test users | Full campaign, press and promotion at once |
| Goal | Learn and fix | Grow |
| Risk | Slower, and competitors can see the app | Problems surface in front of everyone |
| Best for | New products with unproven retention or pricing | Proven products, sequels, apps tied to a fixed date |
The two are stages, not alternatives. A soft launch ends in a hard launch if the numbers hold up, and in more development or a cancelled project if they do not.
When it is worth doing, and when to skip it
Soft launch if you plan to pay for installs, if the app depends on people coming back, or if you are unsure about pricing. Those are the cases where a few weeks of real data changes what you build or spend.
Skip it if the app has to be live for a specific event, if it only works with a large local network of users from day one, or if it is a small utility you can update freely after release. A beta with a few hundred testers may be all the testing that kind of app needs.
What the stores let you restrict
People often mix up three different controls. Only the first is a true soft launch.
By country
Both stores let you choose where an app is available. In App Store Connect you can release in any of the 175 countries or regions where the App Store operates, or only in the ones you pick, and add more later. Google Play Console has the same control per country. This is how soft launches are run: the app is public and in the store, but only in your test markets.
By tester group
Beta programs limit who can install, not where. TestFlight allows up to 10,000 external testers on iOS. Google Play has internal, closed and open testing tracks, and personal developer accounts created after November 13, 2023 must complete Google's testing requirements before they can publish to production at all. Betas are good for finding bugs. They are poor for measuring retention, because testers are not typical users and did not find you through the store.
By percentage
Apple's phased release and Google's staged rollout send a new version to a share of users first. Both apply to updates only. Apple's runs over seven days for users with automatic updates turned on, and Google states that staged rollouts cannot be used when publishing an app for the first time. They protect an existing user base from a bad update. They are not a way to launch a new app quietly.
Choosing test markets
A good test market behaves like your main market and costs less to learn in. Look for three things.
- The same language and similar habits. If the US is the goal, teams commonly test in Canada, Australia or New Zealand, where the listing, the app and the support need no translation.
- Enough users to produce data. A country that sends you ten installs a day will take months to tell you anything.
- Lower advertising costs than the main market, so each lesson is cheaper.
Pick markets to match the question. To test whether the app is stable on cheap Android phones, choose a country where those phones are common. To test whether people will pay, choose one with spending habits close to your target.
Run it as three tests
Game studios have run soft launches this way for years, and it works for any app. Do not move to the next test until the current one passes.
| Test | Question | What you measure | Installs needed |
|---|---|---|---|
| 1. Technical | Does it work on real devices and networks? | Crash rate, load time, bugs in reviews | A few hundred |
| 2. Retention | Do people come back? | First-session activation, then day 1, 7 and 30 retention | A few thousand, in weekly cohorts |
| 3. Monetization | Does a user earn more than they cost? | Trial starts, conversion to paid, revenue per install, cost per install | Enough to see a few hundred purchases |
The order matters because each test is more expensive than the last. There is no sense paying for thousands of installs to measure revenue in an app that crashes, or tuning a paywall for users who leave on day two.
Set your targets before you start
Write down the pass mark for each test, and what you will do if the result is below, near or above it. Deciding in advance stops you explaining away a bad number later. These published medians give you something to anchor on:
- Listing conversion: AppTweak measured an average of 8.56% from product page view to download on the US App Store in 2025, and 16.15% on Google Play. Category averages vary a lot, so look up your own.
- Install to paid subscriber within 35 days: a median 2.1% for apps with a free tier and 10.7% for apps with a hard paywall, in RevenueCat's 2026 data.
- Trial to paid: a median 25.5% for trials of four days or less and 37.4% for five to nine days.
- Revenue per install after 60 days: a median $0.38 with a free tier and $3.09 with a hard paywall.
The target that decides the project is the last one against your cost per install. If an install costs more than it returns in 60 days and your plans are monthly, the app is not ready for a full launch however good the reviews are.
Watch day zero closely in the retention test. About half of all paid conversions happen on the day of install, and so do 55% of cancellations of three-day trials. If new users are not reaching the main action in their first session, fix that before you read anything into day 7.
How to run a soft launch
- Finish the beta first. Fix crashes and obvious usability problems with testers, so the soft launch measures the product and not its bugs.
- Set up analytics and name your events before release. You cannot go back and record what you did not track.
- Prepare the full store listing. Screenshots, description and keywords are part of what you are testing.
- Release in one or two markets and buy a modest, steady flow of installs. A consistent few dozen a day is more useful than one large burst. Buy them from the channel you plan to use at full launch, because users from search ads and users from social video behave differently.
- Review the numbers weekly and change one thing at a time. Ship updates often.
- Compare each new cohort with the one before it. You are looking for a trend, since single weeks are noisy.
- When the targets are met for several cohorts in a row, widen to more countries, then launch fully.
How long it takes and how many users you need
The calendar is set by the metric you care about. Day 30 retention cannot be read until a cohort is 30 days old, and you will want at least two or three cohorts to see whether your changes helped. If day 30 retention is the goal, that makes two to three months the minimum. A game with a complex in-app economy may run far longer.
On volume, the rule is that small samples lie. With 100 installs, three or four people more or less staying changes your retention figure by several points. With a few thousand installs per cohort, three or four people make no visible difference. If your budget only stretches to a few hundred users, use them to judge stability and activation and treat the retention numbers as rough.
What a soft launch costs you
- Time. Competitors in your category can find the app in the test market and copy what works.
- A first impression. People in your test markets meet the unfinished version, and some will not come back for the finished one.
- Ratings, in part. On the App Store the summary rating is specific to each territory, so poor early ratings stay in the test market. You can also reset the rating with a new version, though Apple advises using that sparingly.
App launch checklist
Before moving from soft launch to full release, check each of these.
- Crash rate is low and stable across the devices your users have.
- Retention and revenue per install meet the targets you set at the start.
- The store listing has been tested and localized for the launch markets.
- The rating prompt appears after a successful action, never on first open.
- Analytics, attribution and subscription events are firing correctly.
- Servers have been load-tested at several times the soft launch traffic.
- Support has answers ready for the questions test users asked most.
- Ad campaigns, creative and the launch-day budget are prepared.
What happens after that, from store listing to paid acquisition, is covered in the guide to app marketing. If retention was the weak number, start with the guide to app engagement. And if you have not built anything yet, validating the idea comes before any of this.
FAQs
What is a soft launch?
A soft launch is the release of an app to a limited audience, typically a small number of countries, before its full release. It lets the team measure stability, retention and revenue with real users and fix problems while the audience is small.
What is the difference between a soft launch and a hard launch?
A soft launch is limited and quiet, with the aim of learning. A hard launch is the full release to every target market with marketing behind it, with the aim of growth. Most apps that soft launch go on to a hard launch once the numbers are good enough.
How long should a soft launch last?
Long enough to read the metric you care about across several groups of users. For day 30 retention that means at least two to three months. Simple apps can be shorter, and games with complex economies often take much longer.
Which countries are best for a soft launch?
Ones that resemble your main market in language and behavior but cost less to advertise in. For apps aimed at the US, Canada, Australia and New Zealand are common choices. Choose by what you are testing: technical stability, retention or willingness to pay.
Is a beta test the same as a soft launch?
No. A beta goes to invited testers through TestFlight or a Google Play testing track and is best for finding bugs. A soft launch is a public release in selected markets, so it measures how ordinary users who found the app in the store behave.
Sources
- App Store Connect Help: manage availability and phased release
- Apple Developer: TestFlight and ratings and reviews
- RevenueCat: State of Subscription Apps 2026 (conversion and revenue-per-install medians, for apps using RevenueCat)
- AppTweak: average app conversion rate per category, 2025
- Google Play Console Help: testing tracks, testing requirements for new personal accounts and staged rollouts
Store rules were checked against the pages above on October 5, 2026.