PeopleFun, an award-winning mobile game developer headquartered in Texas, is known for games such as Wordscapes Search, Brinks N Balls, and Blockscapes. It had a heuristic, rule-based recommendation engine for in-game promotions but lacked visibility into player preferences and gaming behavior to convert these promotions into purchases. This was impacting revenues at the business side and leaving gaps in player experience which could be augmented with timely promotions suited to player preferences.
The mobile app monetization model pivots on accurately segmenting online players and predicting the next purchase. This could be challenging without using advanced Artificial Intelligence (AI) or Machine Learning (ML) capabilities.
PeopleFun was already operating on the Google Cloud Platform (GCP). It used Google App Engine to run applications, DataStore to manage data, and BigQuery to analyze large chunks of data from its gaming software. It wanted to collaborate with a strategic partner specializing in developing and implementing AI/ML use cases that can help it enhance its recommendation system with the swathes of player data at its disposal.
PeopleFun turned to Persistent for our AI/ML expertise, data engineering prowess, and our 360-degree, deeply embedded relationship with GCP that gives us the know-how on best-fit AI capabilities in the Google AI stack.
Persistent leverages GCP’s AI capabilities to make PeopleFun’s recommender engines more targeted
The PeopleFun developer team stood at 10, and they needed the data to be cleansed, validated, and ready for analysis to create meaningful promotional campaigns. We used Google’s DataFlow and Datastore to streamline the data workflow from raw data to insights. We leveraged our data engineering and analytics capabilities to create seamless data pipelines that could feed real-time data into ML models, making them adaptive to end-user needs. This also helped the team accommodate changes in the recommender engines based on business requirements, making them even more targeted to unique customer cohorts.
Persistent compiled a personalization stack to power PeopleFun’s recommender engines with Google’s AI offerings. To enable the team to promote different products as per player profiles, we used VertexAI, Google’s ML platform, to train and deploy ML models quickly. This helped the team analyze player data at the edge and segment them based on behavioral traits from activities such as ad watching, purchase history, gameplay, etc. This was crucial to establish the most conducive time to pitch an in-game promotional ad with higher chances of conversion based on the player profile. We also enabled APIs that interacted with the model to offer predictions for players.
BigQuery backed all of this, Google’s fully managed enterprise data warehouse, that allowed the team and executive stakeholders to draw insights from real-time data. This helped the team get more visibility into marketing analytics, enabling them to identify and address notional gaps in their campaigns and achieve objectives faster.
Delivered Higher Conversion and Enhanced Player Experience
With Persistent as a managed services partner, PeopleFun could fully tap into GCP’s suite of AI/ML capabilities, capitalizing on the best-fit solutions to drive its use case. The augmented recommender engines, backed by the GCP-powered personalization stack, can now drive higher conversions of in-game promotions for PeopleFun. This boosts mobile app monetization and helps PeopleFun mine advertising accounts better. We also delivered a superior, all-round customer experience that translates into better app store ratings and more downloads.
Google Cloud technologies and Persistent’s expertise helps us launch new games, add new features, and deliver highly personalized gaming experiences.
Bourke Floyd, VP Engineering, Platform, PeopleFun
With the cutting-edge AI tools in Google Cloud, we delivered a more personalized experience, including in-game promotions, to each player.
Arpit Agrawal, Engineering Partner, AI Services, Persistent Systems
You may also read the detailed case study on Google to learn how PeopleFun uses the extended GCP suite to enhance player experience with low latency and high availability services.
