We have developed Kinrate AI to provide personalized game recommendations to target audiences. Kinrate AI is a hybrid model, which combines both collaborative filtering and content-based approach to generate stellar results.
Kinrate AI has learned to predict both player profiles and latent qualities for digital games. To date, Kinrate AI has learned to portray a total of 12,000 games according to 61 latent gameplay characteristics, which measure six main factors of players’ gameplay preferences: aggression, management, exploration, care-taking, coordination, and problem-solving. We have validated the model in USA, Japan, Korea, Canada, UK, Denmark, and Finland.
“Excellent instant recommendations! I went through the first 30 games in my list, and those games which I haven’t played yet, I’d really like to try”
Solution to the cold-start problem
By teaching AI with player data of more than 48 Million game players worldwide, we have been able to develop a great cross-platform game recommendation engine.
In contrast to established products, our recommender is a plug-and-play model, tailored for game business purposes. Our model can be integrated directly to your tech with our API.
Kinrate search engine finds similar games based on the unique AI-generated latent data. The search engine can be integrated to e.g., your marketplace. By finding the searched title and games similar to it, your customers are likely to spend more time and money on your platform. Try our intelligent similar games search engine.
Understand your player audiences
Enhance engagement by social networkingKinrate player personas can be utilized in developing social network services. Millions of players seek constantly game-pals to play multiplayer games with. By utilizing Kinrate Player Persona approach, you can help players to find their kind of players which enhances retention on your platform, and brings your customer experience to the next level.
Instant personalized recommendations
We can profile your customers by asking them to tell us a couple of their favorite games, by analyzing their online behavior data, or by combining these two approaches. As a result, your users can be given their player type and instant personalized game recommendations. Our existing player type test can be browsed here.
“My player type was 99% Settler, and I totally agree! I had not even heard about some of the recommended games. I wrote a few up, and I will definitely check those later”
“Pretty good match with my profile, I was a 92 % Knight. Game recommendations were also top-notch”
Data-driven competition analyses
You can generate a tailored competition analysis report by accessing our data. A competition analysis report helps you to communicate to e.g. investors why they should consider funding your game. It also helps you to design a marketing strategy to reach your core customers.
Knowledge on emerging trends
By studying our database of 140,000 games and 48M players you can gain knowledge on how players’ gaming preferences change through time. This helps you to make data-driven decisions on what games should be developed and published.
We have combined our player and game data with our extensive player market research data. This makes it possible to predict motivations to play, purchase reasons, game aesthetics preferences, genre preferences etc. of your identified core player-customers. Advanced ML modelling and representative data from USA, UK, Japan, Canada, Korea, Denmark, and Finland
Dr Aki Koponen
Dr Jukka Vahlo
Dr Meri-Tuulia Kaarakainen
Be in contact with us: