Launching a new game or any new feature set in an existing game is a blockbuster moment. There is excitement and anticipation from executives, game developers, and many others. A player’s first experience with a new game is an important moment, especially with so many high-quality games in existence today. Because let’s be honest, first impressions matter. If they have a good experience, then it can be the start of a long-term relationship. On the flip side, a poor experience can cut the relationship short. 

No matter how much quality assurance (QA) work or beta-testing is conducted, the player experience will always be different when the game is released in the wild to real users. If a problem occurs during a launch, it can have a material impact on the revenue, player retention, and brand. Response time affects the player experience. You’ll want to limit exposure to your user base by responding quickly, optimizing outcomes, and resolving bugs. Even if nothing goes wrong, there is always the opportunity to learn from player behavior to improve the experience.

How can teams take a powerful iterative approach to game development? Enter live operations.

What Are “Live Operations?”  

Live operations, or “LiveOps,” refers to any changes made to a game without releasing a brand new version. Experimentation is an important part of LiveOps. This means testing hypotheses about player behavior and game experience, then iterating using the data. Experimentation and LiveOps foster an incredible, player-centric gaming experience. But, how do you know if your team is ready to implement LiveOps? 

There are a couple of questions to consider. Is your team prepared to architect games with modular, independent features? And can they iterate with speed? To meet these objectives, LiveOps teams need to work closely with game development teams. The ability to change features without fear is important, as is data. Typically games that can take advantage of LiveOps have a huge amount of server-side player data recorded to enable personalization and robust data analysis. 

Want to get started with LiveOps, but unsure of how to build out the technology? That’s where feature flags and real-time measurement engines come in. 

Rapid Iteration and Personalization With Feature Flags 

Leading gaming companies modularize game features to enable rapid iteration. They can do this with LiveOps to drive greater monetization and player satisfaction. Pushing large changes game-wide once a week or every few weeks may increase the risk of dissatisfied players, brand damage, and revenue loss. 

Using feature flags with dynamic configurations allows LiveOps and game development teams to iterate multiple times per day. At the same time, they can avoid entire application deployments for bug fixes and performance issues. By pairing rapid iteration with the personalization that feature flags unlock, the game experience can be tailored to a player based on any attribute like location, age, and time played.  

Because feature flags with dynamic configuration can decouple release from deployment, non-technical teammates and LiveOps teams can iterate on existing features without the need for engineering resources. This allows teams to make changes quickly and helps reduce operational bottlenecks. 

Protect the Player Experience With Real-Time Measurement and Feature Flags

If your LiveOps and game development teams can use feature flags with dynamic configurations to iterate multiple times per day, then what will they need to drive informed, future iterations? How can teams quickly understand the impact of these changes?

Oftentimes, teams need to release changes and later rely on downstream analysis. Generally, this is accomplished through a data science or data analytics team. But, these teammates have their own backlog, so measuring outcomes can often take days to weeks. How will teams know that they are iterating in the right direction and not leaving a degraded player experience in production? 

When you integrate data into your feature flagging platform, it can be automatically tied to each feature flag using attribution and statistical engines. This provides event data which can then be used to create metrics. If these metrics are automatically calculated across every change, then data can be democratized to teammates with access to your feature flagging platform. Democratizing data allows teammates to iterate quickly and confidently change that new player gift or lobby matching algorithm. As a result of using feature flags with dynamic configuration, each decision will be backed up with real data. 

Without integrated and democratized data, teams will lack the timely information they need to act with intention. Leaving problems unresolved for any length of time can damage a player’s first impression of a game. 

Suppose data about monetization, engagement, and game performance automatically surface to game development and LiveOps teams for every change. In that case, you will not want to require the additional step of signing into a dashboard to discover that something is going wrong. Who wants to start their day by logging into a dashboard only to find out that the game was experiencing degraded performance or revenue the night before? This is where alerts come in. 

Using alerts, teams are notified when degradation against any business or technical KPI occurs. The alerts provide teams with a link to the exact feature causing the issue. This link gives access to a rollback button that returns the feature to a safe state. With the alerts feature and a rollback button, a team member can identify and resolve an issue within minutes through a simple click. It is no longer necessary to spend an entire night to bug-bash an issue. 

Still not convinced? An experimental approach to game development and iteration is not common, yet is becoming table stakes for the most profitable games. Consider the changes in gaming revenue over the last 50 years: 

Source: 50 Years of Gaming History, By Revenue Stream (1970-2020)

Games that generate the most revenue are the ones that modularize functionality and provide gaming via cloud or mobile. Many of these games are able to stay relevant and top revenue charts due to experimentation that is focused on optimizing the player experience. That’s where iteration comes in. Iteration enables you to provide a personalized and dynamic gaming experience. Without it, you’re left with a static and boring game. And who wants that?

Learn More About Feature Flags and Real-Time Measurement 

The latest and greatest game will always be on the horizon. If you want to stand out, then you need a toolset that helps you create the best player experience. That’s what feature flags with dynamic configuration and measurement can help you accomplish. They will free your team, helping them master the feature delivery lifecycle, operate with greater confidence, iterate faster, and reduce release risk. Game on! 

I hope you enjoyed reading this. You might also enjoy:

  • Feature Delivery Unlocks Speed, Safety, and Business Impact
  • Feature Flags Are Changing Observability As You know It
  • Extending Monitoring from Application Performance to Feature Performance
  • The Surprising Power of Online Experiments  

Follow us on YouTube, Twitter, and LinkedIn for more great content! You should also join the Split Community on Slack.