In 2013, IIT Ropar alumnus Arpit Jain founded GreedyGame to help mobile publishers and advertisers run ads without disrupting the user experience. The company offers a unique Software Development Kit (SDK) that enables app developers to run native ad units customised for their app experiences. The publishers can implement native ads in their apps in a simple and quick manner, ensuring that their revenues are optimised without any policy violations. For advertisers, GreedyGame’s platform ensures that users get the most relevant and high-quality ads. The ads are opt-in, compliant and thus non-intrusive. The result is both brand awareness and high-quality user acquisition for advertisers.
GreedyGame has evolved into India’s largest end-to-end implementation, mediation and optimisation platform for Native Ads on Mobile. The firm works with brands such as Amazon, ONEPLUS, Dream11, MPL, Zee5, RummyCircle, etc. GreedyGame has the world’s first AI-enabled ad unit recommendation and creation engine for implementing native ads.
In an exclusive interview with Analytics India Magazine, Arpit Jain, founder and CEO, GreedyGame, spoke about how the company uses AI to help app and game publishers monetize better with native ad formats.
AIM: How does GreedyGame leverage AI?
Arpit Jain: AI is particularly helpful for sharing and implementing our learnings and optimisation techniques to a larger user set. We use behavioural analysis models to tackle fraudulent or inadvertent clicks to protect our advertiser interests. We also rely heavily on clustering models to target the relevant users for relevant ad campaigns. The app industry typically has high churn rates, so by targeting users with similar interests, we can deliver high Click-Through Rates (CTRs) and also keep the overall budget of the ad campaign low. Also, by matching the user preference attributes for ads, we can generate ad design templates that a user is more likely to engage with when it comes to native ads.
When working with demand partners like Google and InMobi, our forecast models help match the right inventory with the right demand. This ensures that the app and game developers working with us get connected first with the demand sources expected to have a higher return. Since these come along with our SDK X offering, the developers can use them straight out of the box and measure a direct impact.
AIM: How does GreedyGame’s tech stack enhances user experience?
Arpit Jain: Being an ad network, our choice of the tech stack is designed for scale. With a small team, we try to maintain tight SLAs for our APIs to provide a better experience to our users. We have a microservices-based architecture primarily written in Go to achieve a high concurrency. Using a tiered cache mechanism and orchestration tools like Kubernetes helps us to scale in direct proportion to the traffic without breaking the bank. This came in handy when our traffic scaled up 7X during IPL matches, and we could adhere to these strong spike volumes.
Our frontend dashboards and panels are written in ReactJS, which are responsive and can also be used as PWAs. We also have a multi-cloud infrastructure that helps us reduce operational dependency and use the right tool for the right job. Basing ourselves on these newer technologies allows us to be nimble and provide features faster to our users.
AIM: What makes GreedyGame stand out?
Arpit Jain: With our 8+ years of ad tech experience, our tech team has evolved and started curating a set of unique and essential ad monetisation features in our product which makes us stand out from our competition. GreedyGame’s SDK X works as a cross-format mediation, bringing the best of native ads with the regular ads. We have also simplified the integration process for Native ads so that developers have policy-compliant implementations right out of the box.
AIM: How does GreedyGame protect user data?
Arpit Jain: Our systems are designed to exchange the minimum amount of information possible. Since we operate on our internal data and proprietary SDK, the data does not leave our systems. We work with verified partners and exchanges with a privacy policy in place. All internal ids are randomized or masked.
