Predicting new game success

Dunia Games is a lifestyle platform for gamers with over 60 million active monthly users. A subsidiary of Telkomsel, Indonesia’s largest telco, the company provides game downloads and in-game transactional vouchers, hosts tournaments and broadcasts content to its subscribers 24/7. It is now Telkomsel’s largest digital revenue earner.

The challenge

Staying relevant in the gaming market in Asia - or anywhere - is a daunting task. The industry moves fast with trends coming and going at lightning speed. It’s therefore critical to keep up as a business looking to successfully monetise mobile gaming outside of existing services like Apple store, Google Playstore and major publishing groups, who are firmly entrenched in the value chain of the gaming market. A very strong offer was just the beginning for Dunia Games. The company needed to ensure it was publishing and promoting the top trending games on its platform to stay ahead of the competition. With more than 7000 new games entering the Asia market every month (at a paltry 6% success rate) - predicting which new games to publish, and which existing games to invest in, was the key to driving in-game transaction revenue.

Their challenge to us was to provide:

• A list of the 100 hottest games, ranked in the order predicted to succeed in the Indonesian market in the coming 6 months, updated monthly.
• A ranking of top games, already active on the Dunia Games platform, predicted to generate the most in-game revenue over the following 6 months.
• Identification of the best performing game publishers and developers in the region.

“For us as a telco, the secret sauce is understanding what’s going to be the next big trend in gaming then making sure we have built our ecosystem around that. Quantiful has been helping us by using consumer and transactional data to see what’s happening in lead markets like China and Taiwan.”

Crispin Tristram, VP Digital Lifestyle Products, Telkomsel

The Solution

Quantiful’s SaaS forecasting platform, QU, was configured to ingest both download and transactional data from Data.ai (formerly App Annie) for mobile games across six Asian markets (China, Vietnam, Malaysia, Philippines, Thailand and Indonesia). This data, combined with QU’s onboard social and search data for mobile games across the same markets, was used to build predictive models for game performance in Indonesia. The data comprising an average of 12,000 games per period was refreshed bi-weekly and a rolling six-month forecast was generated. In addition to forecasts, key insights into mobile game success factors were generated. Called “QU demand drivers”, these could be used by Dunia Games to identify new trends in gameplay features to help new game design or to support the marketing of games. 

Summary results

Each week QU ingested 600,000 new game data points from Data.ai and 500,000 social and search data points from the target markets, generating a forecast for game success in each market. An interactive QU portal was built for Dunia Games to interrogate the results and view forecast accuracy. QU’s rolling 6 month forecast achieved 90% plus accuracy in predicting the top 100 games successful in the Indonesian market at 8 weeks lag.