Break Pricing in a Small Country
Every price you ship is a guess until a real player pays it. A soft-launch is how you stop guessing in front of your whole market. You open the game in a couple of small, English-speaking, lower-stakes countries — New Zealand, the Philippines, Canada — watch what people actually do with your shop, and fix the dumb parts before the rest of the world sees them. By the time you go global, the prices aren't guesses anymore.
The lesson studios keep relearning: the price that feels right in a spreadsheet is usually wrong in a wallet. Soft-launch is the cheap place to be wrong. A $4.99 pack that looked smart in planning can sit unsold next to a $0.99 one that prints money, and you'd never know without live data from people who aren't your colleagues.
What you're actually testing
It's not just "will they buy." It's which pack, at what price, with what pitch. Soft-launch lets you A/B the shop itself — swap the first-purchase offer, move the whale bait, test a battle pass against a subscription — and read the results in ARPPU and conversion, not opinions. The countries you pick matter less than the discipline: keep it quiet, keep it small, and resist the urge to "fix" things with a hunch instead of a test.
| Thing to test | What a wrong call costs | How soft-launch saves you |
|---|---|---|
| First-price point | dead conversion | find it on 5% of market |
| Pack structure | low ARPPU | iterate before global |
| Sale cadence | trained deal-waiters | tune the rhythm |
| Onboarding paywall | churned newcomers | fix before the flood |
Treat it like a lab, not a soft open
The mistake is treating soft-launch as a relaxed launch. It's a lab. You instrument everything, you run controlled tests, and you make decisions from numbers. A studio that soft-launches without analytics is just delaying the same guesses. The ones that win ship the events, the offers, and the onboarding changes during soft-launch and measure each one, so global launch is a scale-up, not a discovery.
Don't soft-launch forever
There's a trap on the other side: never shipping because the data never feels clean. At some point the marginal test costs more than the risk of going wide. Set a decision rule before you start — a conversion floor, an ARPPU target, a retention number — and when you hit it, launch. Soft-launch is a tool to de-risk, not a place to live.
On a runtime that lets you flip offers and pricing per region without an app-store resubmission, the "lab" can stay open even after global — you keep testing prices in low-stakes markets while the big ones run steady.
