A refreshing reality check on viewable ad units and real time bidding
Article by: Nicholas (“Waken”) Paschal: LinkedIn Profile
What is viewability?:
Generally speaking, an advertiser purchases ads on a CPM (cost per thousand) basis, representing 1,000 ad impressions. An ad impression is quantified by a user visiting a webpage where an ad is placed, regardless of its visibility. Often, ads are loaded at the bottom of extensive blog articles, requiring the viewer to scroll to see the ad. Consequently, more than 65% of ads remain unnoticed by the user, and the advertiser is still charged for that ad impression. To make it simple, if you pay $1 for 1,000 impressions in your advertising campaign, 65 cents of that $1 will be wasted purchasing ineffective ad space. The industry emphasises viewability as a means to resolve this problem. Â An ad is considered viewable when at least 50% of the ad is seen by the page visitor longer than a specified period of time. This process, called CPV (cost per view) media buying, only charges advertisers for ads that have actually been viewed. An ad that goes unseen is the online advertising equivalent of a billboard behind a giant tree. If a website visitor never sees the billboard, then why should you pay for it?
The technical difficulties associated with solving these problems:
Currently, there are several approaches being tested industry-wide.
Using html5’s postMessage API and or fragment id messaging to sync data through iFrames (Both require the publisher to be co-operative.)
Using math to determine the size and location of an ad within the browser window – you can then observe whether the ad was in-view and available to be seen by the website visitor (Browsers can be easily cheated to sidestep this method, and Webkit-based browsers are not compatible.)
Cross browser versions and device compatibility prevents many solutions from becoming heavily adopted.
Creating a hybrid solution is too costly for both the DSP and the advertiser to justify spending the same amount of money on wasted impressions and no viewable filtering. This is, of course, the case if they are a larger brand running campaigns across all devices/browsers. Most DSP’s have a few options and are slowly ramping up technology stacks.
Other ways to look at viewability:
When a user visits a website, their eyes scan and react to focus on the content intended to be viewed. A site that does not place an ad unit properly or has a cluttered environment will perform poorly (see example photo).
Â Marketers need to focus on the quality of the websites that their campaigns are being placed on and the “noticeability” of an ad unit after it has become viewable. This can be accomplished through page-quality information provided through forward thinking 3rd party technology companies integrated with DSP’s.
The closest viewable solution yet:
Modern browsers are progressively trying to increase the speed that sites are viewed. The browser window uses multiple resources to show a visitor in-view content than content outside of the screen to help renderiong, device hardware allocation, scrolling speed, etc.
Â These properties including render frame-rate can be used coinciding with ad tags to measure an ads viewability and trigger analytics when an ad comes in-view.
How marketers should control their visibility:
Report on creative, exchange, domain levels.
View metrics for visibility fold position, duration, surface area of the ad and total page duration.
Establish automatic and manual turn-off modes for exchanges, creative, or domains based on performance.