Evaluating the “right” way to optimize depends on the particular objectives of the customer. For the archetypal publisher, there are concerns about overall monetization and user retention. This means publishers are balancing the overall dollars made from advertising with the number and frequency of users visiting their site. Of course, this grossly simplifies the publishing business, but is nonetheless useful as a point of discussion.
Native demand comes in two forms: direct sold (generally requiring significant customization) and programmatic. The former will likely achieve roughly the scale of custom rich-media campaigns, as they require one-to-one relationships and custom content production. From the publisher’s perspective, this is the most custom and most lucrative campaign that can be sold. Just as many large publishers have healthy businesses from custom rich-media, the same will probably be the case for native. But just like banners and rich media, there is still a huge and burgeoning market for native, and publishers need to make sense of programmatic.
The publisher’s agenda fits in two broad buckets, monetization and traffic. The more native the advertising, the higher the engagement: meaning the advertiser should be willing to pay more. Similarly, the less disruptive an experience is, the less it will have a negative impact on traffic. So when publishers evaluate native, they should evaluate the most native format as it relates to overall yield.
If there were no cost in defaulting to different adservers, publisher would likely be best suited setting up their system to the following path: 1) direct sold native campaigns; 2) programmatically sold native campaigns with highly native experiences; 3) programmatically sold native campaigns with templated experiences. Then, if it makes sense for the ad slot, 4) banner ads.
The strategy above creates a system in which publishers can capture the maximum marginal utility (the native-ness – positively impacting engagement and user experience), and thus the maximum yield, based on availability.
The reality is that there is a cost to defaulting between different advertising servers. So publishers and advertisers need a technology vendor to develop a solution that predicts yield for each step in the process above and automatically creates the highest valued path, normalized for the number of hops that will be required. Once this technology is perfected, both publishers and advertisers will benefit tremendously – pushing forward the state of native.