GumGum, a contextual ad exchange, has partnered with Simplaex’s Rivr, an AI-powered traffic shaping tool to improve its Supply Side Platform’s (SSP) programmatic auction performance. Rivr aided GumGum in responding to recent global events that have led to a surge in supply and steadily declining demand by adjusting ad operations to increase efficiency and performance.
The COVID-19 pandemic continues to threaten economic stability. As a result, spending on advertising has been slashed, resulting in a decrease in DSP spend. In parallel, time spent online increased, leaving SSPs with a surge in supply. What this means for the supply side is more availability that needs to be accounted for, a factor which Rivr has been able to quickly adjust models to in order to adapt and assist GumGum in mitigating the gap for its supply and demand partners.
“By incorporating Rivr into our auction process, our SSP has been able to deliver higher quality traffic to each of its demand partners––driving better performance at lower cost,” said GumGum’s CTO, Ken Weiner, adding, “Using Rivr’s machine learning to tackle more mundane tasks, we’ve been able to direct more of our own machine learning efforts toward our core business priorities.”
The relationship, which began at the start of 2020, has helped GumGum streamline operations. Cutting the amount of bid requests sent out by over 30%, while maintaining 99.5% of existing revenue and facilitating a revenue uplift of 35% per 1M requests. Rivr is able to achieve and improve upon these results using an out-of-the-box traffic-shaping solution that uses machine learning to mould to the server it is working on. By analyzing the data from GumGum’s completed auctions on an audience level, Rivr determines which Demand Side Platforms (DSPs) are most relevant to the inventory. This information is used to train Rivr’s AI-models and dynamically provide actionable suggestions to the servers that run the auctions. The results ensure that demand partners are getting the traffic they are looking for and performance is improving across the board.
“From years of working on both the supply and demand side of the programmatic ecosystem it became clear that improvements needed to be made. Traffic-shaping, with the ability to understand audience targeting, has the power to decrease infrastructure costs and increase performance while at the same time improving the DSP-SSP relationship. This has become especially important as we navigate tumultuous times where there’s been an unpredictable increase in supply and decrease in demand.” said Moti Tal, CTO of Simplaex.
The current realities facing the programmatic ecosystem have further highlighted the technology’s potential in the tech stack of SSPs, Exchanges and Publishers alike. “That’s where we come in,” says Tal “to assist our partners, like GumGum, in improving what they’re working so hard to build.” Rivr is a key tech stack innovation in unpredictable times.
Visit our case study for more information on how Rivr works here