Evolution of Display
The Internet advertising space has always been at the forefront of technology development and scale. From the early years of simply selling banner space hard coded onto a site, more and more innovative ways of using technology have evolved to accommodate the sheer scale and complexity of Internet content and the opportunities enabled through advertising monetization.
In the early years, Publishers and Ad Networks focused on the task of aggregating, harder to sell non-premium inventory to manage and repackage for sale. Advertisers and agencies then focused on the manual task of optimizing this content for best return, largely limited to changing specific sites or ad placements on a weekly basis, based on creative performance. A process that requires significant test budgets and long time lags to gain insights and influence results through optimization.
In 2004, RightMedia launched the first ad exchange, an online marketplace that publishers, ad networks and agencies or advertisers could all use to trade and manage their ‘remnant’ or non-premium inventory. Advertisers and Agencies (the demand side) could define what kind of targeting and ad performance they were after. Publishers and Ad Networks (The Supply Side) could define what kind of pricing / yield they were looking to achieve. The resulting complex marketplace of ads and spaces were then matched using RightMedia’s Yield Management technology to try and get the optimum result for both parties. There were still test budgets required, but the algorithms were able to significantly increase performance over time and automating this task had a dramatic impact on traditional optimization approaches. Seeing this success the big Internet Advertising players began to switch over to this new form of Media Trading. RightMedia was acquired by Yahoo in 2007, Google then followed with an acquisition of the DoubleClick exchange in 2008 and everyone has been moving this way in some form ever since.
The impact of this switch towards media trading through a marketplace and buying optimization based on algorithm decisioning, has unleashed a profusion of new Internet advertising business models and opportunities. The world today see`s SSP’s (Supply Side Platforms). DSP’s (Demand Side Platforms), ATD’s (Agency Trading Desks), AD Exchange and Network models that combine Demand & Supply, as well as solutions focused around targeting audience and data known as DMP’s (Data Management Platforms). Each solution has their own algorithmic requirements geared towards that business’s unique objectives, approach, partnerships and interactions when trading through the marketplace
Perhaps most significantly though, since 2009 we have seen Media Trading evolve toward Real-Time Bidding (RTB) and decisioning; a technique that treats each impression as a unique entity for sale. In a real-time decisioning environment, each advertising impression to be displayed is put out to auction for best potential fit with an advertiser. This ad display opportunity is then sent out to a range of Demand Side platforms with their respective targeting, performance and minimal pricing criteria. The Demand players each have algorithms that evaluate the ads potential to meet their own performance objective – user matching, price performance (eCPM, CTR, CPA) or any other criteria and can choose to bid on it. The RTB Host (SSP or Exchange) uses their own algorithm to evaluate the respective bids and chooses a winner and the ad is displayed.
This whole process happens in milliseconds, tens of billions of times everyday and creates incredible complexity in terms of scale, data-processing, probability modeling and computational decisioning requirements.
Guess who builds the systems and designs the algorithms that do all this? Welcome to IPONWEB!
