Optimization - Chasing a False Positive

  |  March 21, 2011   |  Comments

Is the online display advertising industry throwing "smarter" money away?

The definition of optimization is "an act, process, or methodology of making something (as a design, system, or decision) as fully perfect, functional, or effective as possible; specifically: the mathematical procedures (as finding the maximum of a function) involved in this," according to Merriam-Webster. And a false positive is "relating to or being an individual or a test result that is erroneously classified in a positive category…because of imperfect testing methods or procedures." By combining the two concepts, we paint a fairly accurate and frightening depiction as to what is happening today in the world of online display advertising.

Few dare to admit it, but in today's world of advanced targeting technology, audience buying, advertising exchanges, real-time bidding, machine learning algorithms, and the like, the positive is generally false and therefore the industry is simply throwing smarter money away. How so? Too often the key performance indicator of an online campaign is some form of view-through conversion. Simply put, view-through conversion gives credit to the party responsible for showing the "last ad" prior to an online conversion. While the current advertising technology has brought about some remarkable advances in the ability to accurately target an audience, the advances have also enabled the exploitation of the "last ad wins" model.

If all that matters to a buyer is the fact that you served the last ad on a consumer's machine before conversion, the buyer shouldn't care if the ad is served above the fold, below the fold, to the left, right, or center of the page. The buyer also wouldn't care if the page is brand safe, brand appropriate, contextually relevant to the brand or the consumer audience, or if there is more than one ad or competing brand on the page. Even more appalling is the fact that in this model the buyer shouldn't care if the ad is served 1,000 times to the same user (notice the use of "served" and not "seen"), or that the user is retargeted 1,000 times, or ended up visiting your site as a result of some other paid media.

This could be the result of optimization - optimization to a false positive is more easily achieved now than ever. Imagine if one treated his personal financial investments the same way - it would be akin to sinking 100 percent of your investment portfolio in a Ponzi scheme and continuing to make deposits even when the scam has been exposed.

If we aim to improve the quality and effectiveness of online display advertising, we need to accept that there currently exists no perfect measurement. We have to be bold enough to move away from the detrimental "view through" or "last ad" model, or what I call the "false positive." Given the current global economic and privacy environment, it's of immense importance to optimize the distribution of each ad impression to inventory to make a true marketing impact. We constantly struggle to define true marketing impact and therefore struggle to map a process of refining the delivery of the performing advertising. Perhaps we should pay less attention to the determinant measure and apply more common sense to the act of advertising itself.

Logic dictates that the key to successfully investing in advertising in an oversupplied marketplace is to harness data that is unique to the advertiser or process information that is unknown to competitive buyers, and combine it with the relevance of the Web property where the ad is placed. Consider this pre-optimization, the act of more accurately evaluating where and how advertising will appear, along with its total reach and frequency ahead of purchase. Perhaps the new measure of effectiveness should be old measures such as reach and frequency.

The heightened attention to the use of cookies brought on by the current privacy environment will perhaps aid in a course correction to the tremendous advances in advertising technology optimization. One would hope that these actions would lead to better yield for both the advertiser and publisher while improving the user experience, but this can only happen if buyers stop chasing false positives.



Edward Montes

Edward Montes leads DataXu's sales and marketing activities on a global basis. A respected international leader, Ed has a deep understanding of how advertisers can apply technology to become data-driven and transform their business. More than a decade in the digital advertising ecosystem has given him considerable experience with combining technology, services, partners, and people to deliver optimal solutions to marketing challenges. Prior to joining DataXu, Ed was co-founder and chief executive at Digilant. He also held leadership positions at Havas Media - spearheading several expansion efforts and doubling the agency's digital billings and capabilities during his tenure - as well as Yahoo and TechTarget. Ed is a graduate of Boston University School of Law and the University of Massachusetts, Amherst.

COMMENTSCommenting policy

comments powered by Disqus

Get ClickZ Media newsletters delivered right to your inbox. Subscribe today!



Featured White Papers

2015 Holiday Email Guide

2015 Holiday Email Guide
The holidays are just around the corner. Download this whitepaper to find out how to create successful holiday email campaigns that drive engagement and revenue.

Three Ways to Make Your Big Data More Valuable

Three Ways to Make Your Big Data More Valuable
Big data holds a lot of promise for marketers, but are marketers ready to make the most of it to drive better business decisions and improve ROI? This study looks at the hidden challenges modern marketers face when trying to put big data to use.