Author: Shashank Raj (Business Analyst)
The global gaming market generated revenue of USD 121.7 billion in 2017 and expected to reach USD 180 billion by 2021 at an annual growth rate of around 10%. The CAGR is expected to be greater in the developing economies like India, China, Brazil, and Indonesia among others (as per the Global Games Market Report by Newzoo).
These statistics paint a very encouraging picture for all the gaming companies and budding entrepreneurs in the online gaming segment. But, the development of an online game and bringing it to market requires significant cost. Once an online game is up and running, the maintenance costs are also high.
Game developers have to ensure that player interest must be maintained over prolonged periods of time. The longer players participate in the gaming platform, the more revenue such players will generate. Here, retention is one of the key metrics to predict the revenues and growth of the gaming platforms. It defines the size of the audience, while also informing you about the average time that users spend on the gaming platform, all of which directly influences your revenue.
The growing competition makes both acquisition and retention of users the main concern for online gaming. With a wide range of online games in this dynamic industry, there is constant pressure to retain players.
Leveraging analytics & data science to increase user retention:
An important factor in user retention is to continue providing them new and challenging scenarios to encourage player participation. Due to different demographics, the users have different expectations and it influences the retention. Analyzing the various interactions of users with the game is important to understand motivations behind player decisions, and how to best monetize them. The amount of data being generated through these platforms is huge and can provide necessary insights to actively engage with their user base.
Here we present a two-pronged approach to understand and improve user retention for an online gaming company. Gaming companies can then tweak game mechanics and user incentives to improve user stickiness:
First, Cohort Analysis can help to study user engagement and retention over time across user segments. It is a subset of behavioural analytics that takes the data from an online gaming platform and rather than looking at all users as one unit, it breaks them into related groups for analysis. These related groups, or cohorts, usually share common characteristics or experiences within a defined time-span.
One can segment users by the acquisition channel (marketing campaigns) and acquisition date to analyze the retention rate of each segment. Retention can then be studied across daily, weekly, or monthly time intervals.
Normally gaming companies measure Day 0, Day 7, Day 14 & Day 30 retention metrics. By observing retention across time and segments you can understand at what point intervention is desired and can be most effective. These interventions can range from offering incentives, skipping game levels, free purchases, etc. High early retention can be further analysed deeply to study if game mechanics are of concern and need to be tweaked.
Secondly, to create differentiated & personalized churn intervention for users in the early stages, it is further useful to develop Predictive Churn algorithms that crunch in all available user data to come up with a risk metric for user churn in the early period.
Multiple algorithms like Logistic regression, Decision Trees, Random Forest can be used to predict churn and further complex models can be built. Such models tend to use device attributes, user engagement attributes like average session time, daily opens, levels cleared etc to come up with churn scores.
Users with low retention scores can be given direct incentives to return and play the game each day. This can be achieved by simulation & periodically introducing new content (e.g., artifacts, rules, stories, areas) and ensuring that all content exhibits a degree of persistence to provide a heightened sense for continued engagement.
“One of the companies segmented their users into two groups: those who had been subscribers for fewer than 50 days, and those who had subscribed for 50 to 360 days. They used a statistical model known as logistic regression to estimate the individual effects of selected game features on user retention. Among their findings were that certain games were among those with the highest positive impact on user retention. It was recommended to invest more in customer service on the top “retaining” games and to diagnose pain points of lower-retaining games to identify weaknesses.”
ML and data science applications are becoming an indispensable part of every online gaming company to study user engagement and improve retention, especially in freemium models. Companies need to collect all available user data in one place to really enable analytics teams to perform meaningful analytics. Machine learning can then be applied to create personalized incentives to drive up retention.