MediaGo's Five New Deep Learning Models Accurately Improve Full Funnel Conversion Rates

SAN FRANCISCO, Dec. 2, 2024 /PRNewswire/ -- MediaGo, the deep learning-based intelligent advertising platform, today announced a comprehensive upgrade of its deep learning models. With this upgrade, the models can now accurately assess traffic quality, enable smart bidding for ad placements, and provide accurate predictions at every stage of the marketing funnel, helping advertisers maximize their return on investment (ROI).

Using a deep neural network (DNN) with over one billion parameters, MediaGo is able to process more than 7 million ad requests per second, allowing it to evaluate advertising effects and formulate intelligent bidding strategies in real time.

In recent years, the advertising industry has been plagued by longstanding pain points such as high-volume low-quality traffic and difficulty in completing conversions. To address these common industry challenges, MediaGo has trained deep learning models specifically and introduced five major models that cover the entire marketing conversion chain:

Traffic value assessor: This model can accurately estimate the value of traffic. It allows advertisers to avoid the threat of invalid traffic (IVT), significantly reducing the IVT ratio to less than 10% of the industry average. The model is also able to accurately assess traffic value, eliminating low-quality traffic with poor results, so that bids are only made for the highest-value traffic, improving the effectiveness of media buying.

Attention, interest, and intention prediction models: Together, these three models draw on media data and historical global marketing data to achieve real-time prediction at each stage of the marketing journey, from attracting user attention to arousing user interest and completing the conversion. The attention prediction model provides high-precision estimates of the exposure efficiency of ad placements, increasing the viewable exposure rate by an average of 20%. By accurately judging user interest and conversion intention, the models help advertisers reach users who are more likely to click and convert, increase click-through rates (CTR) by an average of 15% and conversion rates (CVR) by an average of 40%.

Advanced bidding strategy (SmartBid): SmartBid, MediaGo's intelligent bidding product, automatically adjusts bids based on market dynamics and ad performance, offering two modes – Target Cost Per Action (TCPA) and Max Conversion – to meet advertisers'different goals. Data shows that ad campaigns using SmartBid achieve an average 35% increase in return on advertising spend (ROAS).

"MediaGo is committed to maximizing advertisers' ROI using deep learning technology," said Peter Jinfeng Pan, Head of MediaGo. "We believe that by continuously exploring the possibilities of deep learning models, MediaGo can not only help advertisers increase their effectiveness, but also unlock new benefits for our partners and the entire marketing industry."

Photo - https://mma.prnewswire.com/media/2569720/5_Deep_Learning_Models.jpg