[Alarmy] Global Subscription Launch Retrospective

1

Intro

Alarmy is the product of the startup I'm currently working at.

As the name suggests, it's an alarm app, and it currently wakes up about 2 million people every morning. Our main users are in the US and Korea.

As for our premium features, we initially rolled them out to our main target users in the US and Korea back in September 2019.

For about six months after that, we checked if there was a market need (finding PMF) and communicated with our users to make the premium product even stronger.

In Q2 of 2020, we decided to open up these solidified premium features to the whole world. So, last Q2, we focused entirely on this global launch, and the subscription team naturally took this on as a key part of our OKRs.

Expanding globally didn't mean we just went "Ta-da!" and opened it to every country at once. We rolled it out sequentially, country by country.

During that process, we created a "Global Subscription Launch Guide" and used it like a checklist, continuously improving it as we gradually opened up to target countries.

My goal in sharing this is to introduce that very "Global Subscription Launch Guide" we wrote internally. I figure that for a small startup like ours (around 12 people), having something to reference might save you from some of the trial and error when expanding globally.

Recently, many Korean startups have been targeting the global market alongside the domestic one, which also motivated me to write this—it's always good to share experiences. (I firmly believe that my own struggles can save someone else from struggling!)

The guideline I'm sharing is a mix of the following things:

• What a PO needs to handle: Localization priority, defining the scope for usability testing, recruiting testers per country, and pricing policy

• What an engineer needs to handle: Applying translations, Locale specs, payment testing, remote config, and server stability

• What a designer needs to handle: Branding by country and usability testing

• What a marketer needs to handle: Channel-specific localization strategy and ASO strategy

• What CX needs to handle: Customer support strategies for new countries

Before We Dive In: Q2 Results

Before we dive in, if we look at the results first, we gradually opened up globally in stages starting early May,

and you can see the number of premium trials steadily increasing.

Premium Feature Global Launch Guidelines

Now then, let me share the internal guidelines we used during our global expansion.

The guidelines are broadly divided into three sections:

1. Translation

2. Testing

3. Deployment Prep

1. Translation ✍️

1. 앱 로컬라이징은 단순, 문자열 및 콘텐츠 번역보다 신경 써야 할 것이 더 많다.
2. 사용자는 번역된 것처럼 보이는 것보다, 자기 나라 회사에서 만든 것처럼 느껴지게 해줘야 한다. 
3. 같은 의미에 대해서도 나라별 글자 공간 활용이 다르다.
 - 보통 중국어는 영어보다 30% 공간이 덜 필요하다고 함 
 - 우리가 번역하다 보니 스페인어는 영어보다 약 30% 더 길게 나옴
4. Locale Spec에 대해서 챙겨야 한다. (날짜 포맷, 시간 포맷, 통화, 주소 등..)

String Translation

• In-app string translations (splitting the same language by region, like Simplified and Traditional Chinese)

• Content translation (typing missions — quotes for motivation and hard difficulty)

• Translating remote config content (purchase page content)

Checking Locale Specs

Time format

• Alarm time display (it varies case by case, but we decided to use a 24-hour format for Japan)

• TTS format for reading the wake-up time and brief notes

Date format

• Subscription expiration date display

Speaker selection

• Choosing TTS speakers based on the country

Branding Differentiation by Country (branding may differ in some regions)

In some countries, especially China, you really need to put some thought into branding.

For Alarmy, we don't use our original name in Japan and China.

Japan: Okoshitemi (iOS)

China: Shiming Naozhong (Mission Alarm)

(zh-Hans) 使命闹钟 Alarmy — 特大铃声，按时起床
(zh-Hant) 使命鬧鐘 Alarmy — 特大鈴聲，按時起床

2. Testing 🧪

1. 번역만 잘 되었다고 끝은 아니다. 
2. 실제 사용성에는 문제가 없는지 확인해야 한다. 
3. 사용성은 임팩트가 큰 순으로 진행함 (많은 사용자 + 고가치 사용자) 
  - 초기 설치부터 목표 행동까지 이어지는 퍼널에서, 특히 앞쪽부터 확인해 나간다. 
  - 고가치 사용자의 사용성 특히, 프리미엄 기능과 결제에 대해서 확인해 본다.

Translation Testing

• Engineers test the translated strings from step 1 (just checking if the string keys and values match up).

• We recruit native users from each country to test the translated strings (content + pages) from step 1 (for Japan, a site called Konest was incredibly useful).

User Experience Testing

• Free users (the funnel from installation to setting an alarm)

• Potential premium users (the funnel from each purchase page entry point up to the actual purchase)

Payment Testing

• Subscription payment

• Subscription restore

3. Deployment 🌏

1. 나라별 가격 차등 적용하기 (구매력 작은 나라에 대해 배려해야 함) 
2. 배포 직전 설정 확인(서버, 리모트 설정 값) 
3. 배포 이후, 고객 경험 변경에 따른 대비 (CX 대응 협업 요청)
4. 배포 전략 (ASO + 브랜딩)

Applying Differentiated Pricing by Country (this is super obvious, but via each platform's app management page)

• iOS: Appstore Connect

• Android: Play Console

RemoteConfig Setup

• Since we opened subscriptions country by country, we managed the open status per country (+ per platform).

• Remote config for messages to use on the purchase page

Subscription Server (Give advance notice to the Backend team)

• Check if the servers can handle the load (since instances might need to be scaled up if necessary).

Discussing Customer Support with the CX Team

• Discussing strategies for handling changes experienced by users in the new target countries

• Conversion strategies for existing IAP users

• Strategies for cancellation inquiries regarding paid subscriptions after the trial

• Strategies for complaints regarding premium usage

Localization Strategy with the Marketing Team (We put this on hold at the time, just did the bare basics, and pushed the rest to the next quarter)

• ASO strategy by country

• Ad strategy for other marketing channels

Wrapping Up Q2 and Looking Ahead to Q3

Expanding globally might seem incredibly complex to some (crafting localization strategies for each team can feel endless), but it might seem dead simple to others (like, "Don't you just translate it and you're done?").

By breaking the work down into smaller pieces and creating processes, we could make each task more concrete. Because the work was less abstract, we were able to steadily reduce confusion among our members.

Because of this, our team was able to work with a high degree of focus on a few select metrics during Q2.

From a business strategy perspective, we found the PMF for our premium features in key target countries in Q1, and Q2 was a quarter completely focused on user acquisition through global expansion. Actually, we also had goals around premium feature activation, but we only managed to run some A/B testing for that in the final two sprints. In Q3, we're planning a quarter packed with A/B tests to focus much more heavily on premium activation.

I expect we'll go through our fair share of struggles in Q3 as well, but I believe the team will grow thanks to those very struggles.

Every startup's survival strategy is different (case-by-case for each startup), but hearing about various examples really helped us find the right strategy for us.

I think sharing what we've learned is always a good thing, and even just sharing case studies like this is a great start.

The end. Poof!

Let me introduce our company. ❤️

Alarmy