Deep Learning Leads to Ultra-precise Personalization
RTB House introduces new features to recommendations for targeted ads. The solution leverages deep learning (currently the most promising subfield of artificial intelligence-oriented research), in display advertising to achieve more accurate product recommendations, making personalized retargeting up to 50% more efficient.
One of the most common metrics of a good ad campaign is its click-through rate (CTR), the ratio of clicks on a banner against the number of total users who view the ad. The goal is simple – draw in potential buyers by offering the most click-worthy creatives.
Retargeting and personalization technologies have shaped the way marketers offer the “perfect” creative, by targeting users with a unique offer based on their browsing history. Now, deep learning methods will enable a new generation of ad personalization.
RTB House has developed an innovative recommendation mechanism that uses deep learning to find patterns in decision-making. It can predict a product range that is most likely to be attractive to users, based on their habits, behaviour and other metrics. Behind the scenes, this method uses a mathematical model inspired by the biological neurons in our brains (a so-called artificial neural network), which makes it possible to derive more reliable, richer, machine-interpretable descriptions of customer’s buying preferences from data.
Bartlomiej Romanski, Chief Technology Officer RTB House, notes that what was once a sci-fi concept of predictive intelligence is now a part of everyday reality: ”Deep learning is helping us to unlock the ‘black box’ of big data, enhance our offering to clients and grow our business. By elevating the accuracy of product selection we can make our personalized ads even more precisely targeted, bringing highly optimized ROI for clients, while also giving them business insights to use their budget more efficiently.”
Deep learning consumes an enormous amount of computing power. To develop and implement deep learning algorithms successfully requires a computing platform optimised for this application. RTB House has turned to NVIDIA’s GPU (graphics processing unit) accelerated deep learning compute model for training and inference support.
Serge Lemonde, NVIDIA Deep Learning Startups Business Director EMEA and India explains: “Deep learning is a powerful approach to AI because it allows software to learn from data. Machines are now able to learn at a speed, accuracy and scale that is enabling every industry to tap into the power of AI. RTB House is taking a leading role in bringing this technology to advertising, uncovering business insights in its customers’ data to offer more relevant content to consumers.”
RTB House is one of few companies in the world that managed to develop and implement its own technology for purchasing advertisements in the RTB model (real-time bidding). The company operates worldwide and runs more than 1000 unique campaigns for global brands in 40 markets across Europe, Latin America, Asia and Pacific, Middle East and Africa.
RTB House is a member of NVIDIA Inception Program dedicated to companies who are revolutionizing industries with advances in AI and data science. It is RTB House’s third implementation of deep learning based algorithms after a model for accurate conversion probability and conversion value estimation launched a few months ago.