Featured Blog | This community-written post highlights the best of what the game industry has to offer. Read more like it on the Game Developer Blogs or learn how to Submit Your Own Blog Post
Drowning in the data lake? Get insights into Square Enix Montréal’s BI team and processes, and how they help our teams successfully develop, launch, and operate games.
--- The BI Iceberg: A Key to Mobile Game Success ---
In today’s video game industry, the search for quality, usable data is the new gold rush.
Who hasn’t been hearing non-stop about machine learning, data science, data pipelines, and Artificial Intelligence?
Business Intelligence (BI) plays an increasingly crucial role in companies of all fields since it brings together the processes, expertise, and technology required to leverage data.
In a nutshell, BI enables companies to make strategic and business decisions.
How can you avoid drowning in the data lake? By thinking everything through from the onset of your production and anticipate what is to come.
Mobile games are complex systems, and the flow of data is considerable since it is collected in real-time from multiple sources.
Here are insights into Square Enix Montréal’s BI team and processes, and how they help our game teams successfully develop, launch, and operate games.
A lot happens before the BI team can deliver results and insights to the production teams.
Often, only the tip of the iceberg is visible through outputs such as a result value, a p-value for a test, a model or dashboards.
The starting point of a BI process is usually dealing with a data lake.
A data lake is a storage repository that receives data from different sources. It can also begin with a data warehouse where we store data in a more organizational architecture with files, folders, tables, etc...
The next step is integrating all this data through a set of pipelines that will feed a Cloud Computing Platform (CCP).
What comes next is enriching our knowledge of our players using a CDP (Customer Data Platform). This will enable us, for example, to estimate a player’s Lifetime Value (LTV) and help the LiveOps team make decisions.
The CCP will also plug in several tools: visualization, reporting, and analytical tools. That will help the data analysts performing their analysis and then impact on product decisions.
Square Enix Montréal’s BI team is composed of more over 10 specialists who work together with QA, back-end and tech, production, and user researchers:
Data Engineer/Developer
BI Programmers
Data Analysts in production or in the Central Team
Leads
Director
The strength of Square Enix Montréal BI team is that the roles and skills of team members are not independent; they overlap.
Each specialist has their area of expertise, and most importantly, general expertise that allows each of us to understand and support the global process.
The Data Engineers/Developers can be described as the data plumbers. They are the first to deal with the initial data soup. They build the relevant pipes (Data Pipelines) to organize this soup and redirect it into the CDP following a structured architecture.
The BI programmers are the data catalysts. They speed up the digestion of data. They act as the interface between teams on technical and analytical topics. Their goal is transforming the datasets to extract meaningful information, present information in a more digestible way, and implement best practices to operationalize the data.
The Data Analysts are the data artists. Their primary goal is finding relevant, creative, and innovative ways to answer the game teams’ questions and understanding the player behaviors. Indeed, there is a difference between intended and actual player behaviors. Analytics allows us to see difference.
Also, User Research expertise allows us to better understand the intended behavior. The value of the Data Analyst's expertise is showing what players are actually doing so that the stakeholders and dev teams can adjust the game’s features. However, the prerequisites upstream are building good tracking specifications and maintaining a high level of data quality with the help of the QA team.
The Data Scientists are the Sherlock Holmes of data. They extract insights from raw data from disconnected sources. They also work with more complex modeling such as predictive modeling or machine learning procedures. They can be called to work with the Data Engineers and BI programmers to build machine learning pipelines.
Other roles are also key to a successful BI collaboration including the legal team since they are responsible for handling data compliance and data security in accordance with the Data Protection Regulations worldwide.
At Square Enix Montréal, we have committed to involving BI in the development process at every step. We achieve this goal thanks to the expertise of the different teams and the constant work of stakeholders and leads to keep BI implicated in all the processes.
In the past, there was a common misconception in many industries that the contribution of BI was limited to "a simple extraction" of data or a p-value.
Once you become aware of the entire BI process and the many people and skillsets involved, you understand that a question asked at the tip of the iceberg requires that the entire BI chain (i.e., the submerged part of the iceberg) be solid.
The more people outside a BI team understand that there is a massive iceberg of roles, skills, and processes submerged under water, the easier it will be for companies to reap the benefits.
In innovative and evolving fields such as Data Tech and Data Science, it is important to always keep an eye on what's new and what's happening.
Training is a good starting point to keep up. Attending conferences and events such as meetups are great ways to socialize with colleagues in the field (post-pandemic) and to realize that we are all on the same iceberg..heu...boat.
Thanks to my colleagues for their contributions to this piece: Jessie Chen, Charles Desmonty
Read more about:
Featured BlogsYou May Also Like
