AppNexus, the world’s leading independent advertising technology company, today announced the launch of the AppNexus Programmable Platform (APP), the industry’s first programmable Demand Side Platform (DSP).

The group claims the new product harnesses advanced machine learning to help professional traders execute “highly-customised and complex strategies” through a user interface.

“Most demand side platforms today don’t fundamentally differ from the DSPs that were built during the first decade of programmatic, and that’s obviously a problem,” said Brian O’Kelley, CEO, AppNexus.

“The next generation of DSPs should leverage machine learning to more efficiently manage an ever-larger number of campaign variables, minimise manual intervention, and deliver superior optimisation tied to marketer KPIs. With APP, the world’s first and only programmable DSP, traders can spend less time on setup and delivery and more time strategising on the intuitive, creative aspects of advertising that change a person’s perception of a brand.”

“We’ve reinvented the DSP”

The launch of APP represents the culmination of a three-year R&D cycle involving over 200 dedicated buy-side engineers, data scientists, and product developers. It also signals AppNexus’ aggressive 2018 go-to-market effort with trade desks, agencies, and marketers.

“If we seemed quiet on the buy-side, that’s because we were,” O’Kelley continued.

“To build a next-generation demand platform rooted in machine learning and data science takes time. In the same way that we reinvented the ad server, we’ve reinvented the DSP.”

APP also challenges what Appnexus calls “outdated paradigms” of cookie-based targeting and impression-based buying. The new product provides deterministic cross-device capabilities for people-based targeting at no additional charge and enables a “viewable-only” buying mode.

It also provides built-in supply path optimisation, which charts the most efficient path to supply when traders are operating Open RTB – a key capability in an environment where header bidding has created a duplicative and noisier supply landscape.

Automatic bid price optimisation harnesses machine learning to shade bids when traders are operating in first-price auction environments, thereby allowing buyers to win auctions at the right market price. APP also deploys a combination of machine learning and human intervention to prevent fraud and ensure brand safety.

“We were determined to reinvent the DSP, and we built a new product from the ground up that transforms the role of the trader. There aren’t enough hours in the day to perform the same calibrations and interventions manually, and APP allows the professional trader to get closer to strategy,” said O’Kelley.