On Wednesday, [x+1] announced an upgrade to its Predictive Optimization Engine (POE) platform that allows agencies and marketers to automatically optimize online ad impression bids. The platform now provides impression-by-impression bidding on exchange and non-exchange inventory through an intuitive user interface [UI]. It becomes available June 15.

Think of it this way. Someone clicks on a link. And by the time it takes for the browser to connect to the Web site, POE matches the data in the cookie to analyze and verify the characteristics of the data.

POE must look up the model for the client to determine the ad buy. It generates the bid and pushes the information back to the exchange. Chances are the same ad gets multiple bids simultaneously. The platform also must communicate to the ad server to pull the correct content and serve up the advertisement. Whew.

Nardone explains the 60 milliseconds is the service-level requirement from ad exchanges, a requirement that gives them enough time to serve ad without causing a delay in rendering the Web page. It's an incredible technical challenge, especially when there are thousands of transactions per second.

[x+1] reengineered the upgrade version of POE from the ground up to maximize ad performance and optimization in RTB environments, such as major ad exchanges like Google and AdECN and publisher optimizers like Pubmatic and AdMeld. Using both contextual and audience data, it determines in real time the optimal price to bid for each impression.

Nardone says POE has a learning curve for new clients, but engineers have developed something he calls "pre-targeting," which shortens that time. Pixels are put on the client's Web site before the campaign goes live to develop the targeting clusters. That gives the engine a head start to optimize ad targeting.

Last week, [x+1] announced upgrades to its demand-side platform (DSP). Campaign performance metrics, reporting and analytics are now consolidated across inventory sources into one view in the dashboard, and optimization and context-management tools are available through a UI.

The reporting and analytics platform has been upgraded, too. Instead of having to manage campaign data from various purchase sources-exchange-based, non-exchange-based, RTB, non-RTB-in separate spreadsheets, a campaign manager can access all the data through the [x+1] UI, gaining one consolidated view of all digital media campaign activity. The new version also becomes available next month.

This commentary is insightful. I recommend it to others.