An Online-Advertising Killer App?

One of online advertising’s longstanding promises is the ability to optimize campaigns, allowing planners to increase media efficiency and effectiveness after a campaign has launched.

Optimization gives online media an advantage over other media vehicles, such as television and print, which offer little flexibility after the buy.

In theory, the process is simple. Planners place various creative units across multiple sites. Once the campaign is live, they monitor performance by site and creative. The weakest units are pulled, and the best performing sites (or site areas) are given more media weight, thereby boosting effectiveness.

But, in the real world, obstacles make optimization difficult. Trafficking through third-party ad servers can be a tremendous hassle. Navigating through layers of review and permission can slow things down. And when a plan includes dozens of sites, making even simple modifications to a live campaign can be a Byzantine nightmare.

An interesting company called Paramark is trying to make the optimization process easier. Its product, AdPilot, is an “auto-optimization” engine that uses algorithms to continually optimize based on campaign objectives.

AdPilot monitors live campaigns and can automatically weed out poorly performing creative units and increase media weight where a campaign is most effective. The system, which plugs into ad-serving systems, promises to take the guesswork and hassle out of media optimization.

The product has proven successful in a number of case studies. According to Sanjay Ranka, Paramark’s chief scientist and cofounder, the product recently increased a campaign’s overall registration rate by 113 percent during a campaign running on a business portal. While many of the campaign’s banners generated similar click-through rates, a subset produced the highest registration rates. AdPilot automatically recognized these banners and increased their frequency in the creative rotations.

Smart direct marketers test different combinations of ad attributes, such as offer, message, and design, to isolate the most effective drivers of response. But this can involve hundreds of test cells, making optimization extremely complicated. AdPilot does it automatically, selecting the most effective ad attributes for direct response and optimizing based on the right mix.

If you are an advertiser or agency and your objective is direct response, AdPilot is worth looking at. Paramark offers an efficient way to take the drudgery out of the lives of media planners, letting them focus on more strategic issues. That makes it more likely that media plans are well put together in the first place.

But it is important to keep in mind that the tool, in its current configuration, reinforces a reliance on direct response metrics that the industry is trying to transcend. When AdPilot incorporates metrics that reflect a broader set of advertiser objectives, it will have the potential for becoming a killer application for online advertising.