Deconstructor of Fun

View Original

How Wooga's Switchcraft Aced the Soft Launch!

Many of you know how to develop a game and are experienced in running live operations. However, only a few truly know how to launch a modern mobile game. In the rapidly evolving games market, best practices on how to successfully launch your game can change just as fast. 

To shed some light on the most recent best practices, I had the privilege of talking to Annelie Biernat (Game Director at Wooga) and Aaron Hiscox (Business Development Manager at Google Play).  

Annelie and her team recently released Switchcraft after going through a challenging soft launch. In our conversation below, we discuss key learnings from her experience and compare them to wider best practices observed by the Google Play team. 

Annelie Biernat (Game Director at Wooga) and Aaron Hiscox (Business Development Manager at Google Play) shared their insights and learnings on how to successfully complete a soft launch in today's market.

See this content in the original post

Please enjoy the podcast above and check out the shortened transcribed version below. If you want to make sure you get all the details, I suggest listening to the full podcast.

You can also learn more about acing your launch and optimizing your game’s success by checking out Google’s guide on releasing with confidence and taking this preparing for launch quiz.


Why are developers gravitating towards Google Play during the soft launch/testing phase of their games? 

Aaron:

Google Play offers developers a wide range of soft launch options in Play Console, such as closed testing, open testing, and pre-registration. This flexibility means that developers can choose the soft launch strategy that suits their game and what they’re trying to achieve. 

For example, maybe you’re working on a well-known IP and need to keep your soft launch under wraps. In this case, you’d run a closed test. When you’re ready to start scaling your soft launch, you can run an open test and make your game visible to players on the Play Store with the caveat that your game is still in development and might not be fully ready.

The Play Console also provides the option to run store listing experiments that allow you to test and find the best icon, screenshots, videos, and descriptions for your game on Google Play. In addition, the Play team offers helpful peer benchmarking to give developers a sense of what they should be targeting and where there are opportunities to improve.

Annelie, why did Wooga choose Android over other platforms?

Annelie:

While the services offered by Google played a big role, we also did a market analysis looking at recently launched games and the share of both users and revenue was very compelling on Android.

We soft-launched our game, Switchcraft, on Android and iOS at the same time. When we soft-launched it was during the time of massive uncertainty stemming from Apple’s privacy changes. It was unclear how to do performance marketing effectively when the changes took place and how to measure campaigns. Overall we were unsure if our learnings during the soft-launch would be transferable to our global launch. This uncertainty kept the investment into marketing at a minimum.

On Google Play, we initially faced some technical issues with open beta and couldn’t get the campaigns running. Once in touch with Google, we received technical support and other great feedback. We even set up a call with the Play representatives and our game team to talk about how the user experience can be improved in our game. 

We also benefited a lot from the benchmarks Aaron mentioned. They helped us to set targets and focus on KPIs where we were lacking compared to the competition. 

How do you approach building your go-to-market plans? What are the different phases before you enter soft-launch and during soft-launch? What are you seeking to test in those phases and what are the durations, exit criteria, and budgets for each of the phases?

Annelie:

Before we committed to fully building and entering production, we started years ago with a validation release. There wasn’t really a game in the market that used our game loop, so we wanted to validate a few key hypotheses: 

  • Is there an audience for a game like ours?

  • Does the game loop work? 

  • Does the chosen theme match players' expectations? 

Scope-wise we focused on key features and content as well as two points in the game that would trigger player surveys. We used the information from the survey but also arranged calls with a few players to get an even better understanding of the audience through qualitative feedback.

In 2020 we started with a technical soft-launch in Indonesia and the Philippines. Our goal was to make sure that the game is technically stable. In this phase, we measured loading times, funnel, and level performances all while being focused on fixing bugs.

We soft-launched in the UK, Canada, Australia, which is not typical. The reason was that the narrative plays a huge role in Switchcraft. And since the narrative in soft-launch was English only, we had to release the game in English-speaking countries to properly test the game. 

Besides getting the feature scope in place, our main focus was finding scalable marketing channels and networks. We tried to push boundaries as much as possible to understand the limitations and weaknesses, while also optimizing the funnels. We got frequent benchmark updates from Google to see how we stood against the market standards.

The global launch happened at the point when we were satisfied with the scope and also stopped learning anything new in the markets we had entered. After the call to launch the game globally, we had a dress rehearsal phase where we put everything in place for global launch.

Each phase had defined KPI targets and a defined schedule to reach them. The technical soft-launch was planned for 2 months but took 3 months because we had an issue that needed to be fixed. The soft launch was planned for 3 months but took 6 months because the first attempt failed after only one and a half months. What happened was that we quickly discovered that our FTUE (first-time-user-flow) was not performing. So we had to stop UA (user acquisition) to focus on rebuilding our FTUE. 

When we restarted, things looked much better. Our early retention was where it should be, though we didn’t achieve our long-term retention targets. This forced us to focus on D30 retention for another couple of months before reaching the KPIs needed to commit to a global launch.

Besides the KPIs a big question was always about the team. We asked ourselves if we had a sustainable set-up and if we had the key skills to continue to push the product to the next level. At this point our team was made up of 35 to 40 people and we had already invested more in UA and marketing during the soft launch compared to the investment into the team and overhead.

After the technical soft launch, Switchcraft was forced to take a step back and focus on reworking the FTUE (first-time-user-flow). Once that was fixed, the team worked to reach their long-term retention goals together with marketability targets. These two together set the foundation for the business case and the team entered the final phase of soft launch called “dress rehearsal”. In this phase it was all about buttoning down for the big launch.

What are some of the latest insights in goal setting and performance measurement? For example, what does great engagement look like? What are benchmarks for D1, D7, and D30 retention rates and what type of targets should developers have for monetization?

Aaron:

The key metric that developers continue to put first is retention - if the game retains well, then it’s easier to improve other core-KPIs later. A lot of Western developers still view the US as their main target market, however, we’ve seen top developers do incredibly well by prioritizing markets in Central and Eastern Europe, South America, and Southeast Asia. 

Since benchmarks/target metrics vary so drastically by game genre, I’d suggest looking at the Reach and device insights in Play Console to build out a peer set for your ideal game.

How do you build a media plan for soft launch? What are the criteria for networks, how do you set a budget, what are the sample sizes, what determines the length, and how do you choose markets?

Annelie:

I am no marketing expert. I see myself more as a partner to our amazing publishing team here at Wooga. During our soft launch, we tested the key partners first: Google and Meta with different kinds of campaigns.

We were aiming for a certain size in daily cohorts (in our case the target was 300 installs) and that determined the budgets. We would increase if needed or decrease if no players were needed. We also tested updating campaigns over time and we tested over 100 ad creatives to find out which messages resonate the best and brought the best kinds of players. We also conducted several qualitative audience studies to find the right messaging. 

The campaign length depended on product needs but also the time that campaigns needed to start performing. As you know, some networks take up to a month to have campaigns fully running at a scale and budget. 

Later in the process, we also tested other channels that we identified as important to our key audience like Instagram & TikTok. Since we only planned to be localized to all the major languages by global launch, we chose markets by language. 

How do you advise your studios to approach the soft launch phase from a media perspective?

Aaron:

Testing ad creatives is a critical part of soft launch. So many games succeed or fail based on the strength of their user acquisition and marketing. Simply put, if your marketability is poor, acquiring players will be very expensive. This leads to incredibly high KPI demands for the game, which are often impossible to overcome.

Even at soft launch, App campaigns can help to build up your core user base and generate a rush of installs before scaling your reach and optimizing for your revenue goals. CPI (cost per install) and ROAS (return on ad spend) are common metrics used by developers to determine how well the game is scaling within the first 30 days. 

My advice is to test ad creatives in as many markets as you can before launch. You never really know where your game is going to take off and established markets can be so competitive, so it’s great to gain traction elsewhere.

Since you can tailor your Play Store listing by market, it can also help to test localized creatives such as changing the color palette, tone, incorporating cultural moments and trends, etc.

A strong pre-launch strategy is key to setting up your game for success and getting a headstart towards driving installs and revenue within the first month. App campaigns for pre-registration are a great way to drive installs within the first week to tap into the golden cohort at launch. 

How early do you start testing ad creatives? How much time do you invest into understanding your target audiences?

Annelie:

The truth is, ad creatives are never tested early enough. I wasn’t part of the concept phase of the project, but I think creatives could have been used as concept refinement early on to avoid some of the detours we took during the production of the game.

We tested new ad creatives in every campaign besides the best working old creatives. We used surveys to get a grasp of which style resonated with the player on various platforms because we didn’t want to rely on a singular source of information. During the soft launch, it was probably 30% of the product marketing managers' time that went into audience research in addition to all the research that was done on the product side.

In what ways does Play help developers with creatives and app store optimization?

Aaron:

Play helps developers with this through custom store listings. They allow you to tailor your app’s store listing to appeal to specific user segments in the countries that you choose.

I also recommend running A/B tests in the Play Console via store listing experiments to optimize your creatives and messaging. My old colleagues at Pocket Gems would use this functionality all the time to get empirical data on what combination of assets was most effective in driving conversion rates on the Play store listing page.

We regularly review and update our policies and recommendations for the Play Store. They’re always worth reviewing before you embark on a project to update your store listing. For example, we advise developers to include at least three landscape screenshots to take advantage of the refreshed store look and feel. We also advise developers to limit text overlay and try to use actual gameplay in videos and screenshots, as opposed to cinematics - the logic here is that gamers WANT to know what game they’re going to be playing. My personal recommendation is also to always look at what the top games are doing!

In addition, we recommend integrating the Google Analytics for Firebase SDK at the start. It is a powerful tool that can help optimize ad campaigns performance and reduce campaign ramp-up time. In addition, the SDK provides all the first-party data that you need to understand, segment, and engage players in-app.

On Play, we allow developers to run ads for games in open test tracks and during pre-registration - so this is a great way to test and see what resonates while minimizing risk.

What are some features that require significant investment but don’t actually matter that much in the soft launch? For example, is it worth the investment in optimizing the FTUE or adding offline gameplay ? Also, can you highlight some of the features that give the most bang for the buck?

Annelie:

Supporting offline gameplay sounds simple but is actually very expensive. It adds complexity to every feature and really limits what we can do with the extensive content both on the levels and story side. It's been a staple feature for Match-3 games for a long time but we decided to not do it as we didn’t think it makes a difference for the business case as there are enough successful examples that don’t support it.

If social is not at the core of your product, it is also a potentially very expensive feature set to even implement a lite social experience, even if all others in the market have it by now.

Optimizing the FTUE is definitely worth it - it is important to invest in a modular and adjustable version. Our first iteration took us almost 1.5 months, our second 3 weeks, and the third just one week because we put systems in place that can easily be adjusted. We could see that increase in FTUE completion directly translated to better D1 retention and a higher D7 curve. If done right, optimizing the FTUE can improve the bottom line. 

In the past, I have also worked on a hidden object game where changing the FTUE did nothing because those changes just couldn’t fix the core issues. Some people simply don’t like hidden object games and there’s no FTUE that can change that.

During Switchcraft’s soft launch, we tried to focus on the key experience of playing match-3, how players experienced the story and reducing the loading times. To us, those elements served as the stable foundation of the game.

Right after the global launch, we implemented a simple milestone event that didn’t take long to develop but that generated great results. We could have of course worked on it much earlier, but I think we wouldn’t have achieved the same results without a solid foundation.

When do you know you’re ready for global launch? What are the final steps?

Annelie:

If you can put your KPIs into a believable business case forecast you know you are ready to globally launch. It boils down to a healthy LTV (player lifetime value) > CPI (cost per install) and campaigns that don’t run dry soon after they are kicked off.

Our final steps during the dress rehearsal stage were:

  • Talking to our platform partners and pitching for featuring opportunities

  • Producing the launch trailer(s)

  • Preparing company PR

  • Conducting quality control on the translations

  • Localizing the storefronts

  • Activating pre-registration on Google Play

  • Starting pre-registration campaigns

  • Budgeting allocations to campaigns, networks, and countries 

  • Creating stable build weeks in advance to make sure that no technical bugs will impact the global launch

  • Managing post-launch expectations

We decided to launch globally because we believed that doing so would allow us to continue to learn at a global scale and in the US, our main market. In the end, a global launch is just another milestone in this marathon to build and maintain a viable business case.


You can also learn more about acing your launch and optimizing your game’s success by checking out Google’s guide on releasing with confidence and taking this preparing for launch quiz.