//SSP Smaato adds machine learning to direct inventory to DSPs

SSP Smaato adds machine learning to direct inventory to DSPs

 

 

Platform Offer Side Smaato announced Tuesday the addition of machine learning to its Automatic Traffic Control (ATC) feature, a step that he says can "dramatically" reduce platform traffic overload on demand. (DSP) and lead to better prices for publishers.

What does this mean? For about a year, the San Francisco-based company has been using its air traffic control to better handle the ad requests in the inventory of more than 90,000 application publishers it represents.

Suppose a location in an app in Publisher A calls an ad in the Smaato exchange. The exchange then sends the request to one of the 260+ DSPs in its network.

However, Freddy Friedman, Product Manager, told me that the DSPs were inundated with applications for applications and other inventory, due to the fact that they were inundated. increase in the traffic generated by the header auction and other factors.

Some DSPs undertake their own efforts to "strangle" the flood by arbitrarily filtering certain advertising demands. This allows these DSPs to better manage the bids received for remaining inventory requests, but it also means that some stocks are set aside.

Why machine learning makes the process easier. Prior to today, Smaato's ATC was helping to direct ad requests to selected PSDs instead of spreading them broadly, using manually-defined rules on spam rates. CPM, clickthrough rates, territory, types of ads, time and other features of each DSP. . The same ad request is often sent to multiple DSPs.

Now, added Friedman, the addition of machine learning makes it possible to modify the flow of ad requests sent to DSPs in real time, depending on the current behavior of each call for tenders. Smaato says its air traffic control can now reduce unwanted offers by 72 percent, thanks to better DSP targeting.

This makes it possible to decide which inventory should be permanently changed on the fly, instead of following static rules.

He added that while he assumed that other exchanges similarly directed advertisements for specific PSDs, he was not aware of any third-party advertisements indicating that they were using machine learning.

Why this is important for traders. Directing inventory ad requests to the most favorable bidders is the primary goal of digital publishers. As a result, Smaato's enhanced ATC – if it works as expected – could help publishers sell the most advantageous application inventory at the best price. Fewer more interesting inventory opportunities can result in higher offers and help publishers to place all of their inventory.

Friedman says that intelligent boating advertising requests by ATC also reduces the time and costs for DSPs, a savings that could then be passed on to advertisers. He also says that the new system helps advertisers by matching their offers with the most relevant inventory.

This story was first published on MarTech Today. For more information on marketing technology, click here.

About the author

Barry Levine covers marketing technology for Third Door Media. Previously, he covered this space as a senior editor for VentureBeat, and he wrote on these technical topics, among others, for publications such as CMSWire and NewsFactor. He founded and managed the website / unit of PBS Thirteen / WNET; worked as a Senior Producer / Writer Online for Viacom; created a successful interactive game, PLAY IT BY EAR: The first CD game; founded and directed an independent film, CENTER SCREEN, based at Harvard and M.I.T .; and served for five years as a consultant to the M.I.T. Media Lab. You can find it on LinkedIn and Twitter on xBarryLevine.