Making Sense of Online Video Advertising
With a mix of video formats, delivery platforms, and ad types, online video advertising is anything but simple. Learn where the industry is now and where it's going in the near future.
It’s not as complicated as 3D chess -- you remember, the one Mr. Spock used to play in front of a staring, incredulous Dr. McCoy -- well, it’s not quite as complicated. But video advertising to multiple devices isn’t exactly straightforward at the moment.
As Tim Siglin wrote in the 2013 Streaming Media Industry Sourcebook, “[T]he majority of online content viewing is now viewed on portable and mobile devices.” But can we get a simple and clear view of what’s possible, or do we really need Spock’s brain to understand the options?
As Luke Gaydon, VP of Operations, EMEA at Brightcove, puts it, “Publishers aren’t always aware of the inconsistencies across device platforms and operating systems. These inconsistencies can prevent smooth video playback and block reliable ad delivery and analytics capture, resulting in poor user experiences and ineffective content monetisation.”
Poor user experiences. Blocked ad delivery. Ineffective content monetisation. It just isn’t logical!
The Original Solution
Video advertising has traditionally focused on the mature Flash environment, where interoperability is well-developed.
On a publisher’s website is an embedded Flash player. When the player finishes loading, it calls out to its assigned ad server URL and asks for a pre-roll advert -- a video ad that should be shown before the main feature video. The ad server works out which of its many ad rules it should follow (what is the publisher’s website; where is the viewer geographically; which campaigns could be applied, etc.) and returns a single video ad URL and some tracking URLs. The Flash player plays back the pre-roll ad and reports on whether a user played the ad all the way through, and then it seamlessly continues playback of the main feature (Figure 1). With me so far? Good.
Figure 1. The ad insertion process: 1. Call—player asks ad server for an ad; 2. Response—ad server responds with the ad details (URL, tracking URLs, etc.); 3. Preroll—player pulls in the preroll from the video server hosting the preroll ad; 4. Featured content—player pulls in the featured content from the video ad server hosting the main feature.
It’s all kicked off by the “ad call” -- the player calling the ad server.
So a pre-roll and a main featured video is just a very simple playlist. Before we move on, let’s just be sure we understand how an advertiser can measure the effectiveness of its ads, and how an ad server does its best to target the ads to the viewer.
Tracking and Targeting
Video advertising effectiveness can be measured in several ways. Was your pre-roll shown to a viewer? Did your viewer watch the ad partway or all of the way through? Did your viewer click on the ad and get taken to a webpage with further information? All are important questions to an advertiser concerned about return on investment.
To find the answers, advertisers provide a variety of tracking pixel URLs, for each parameter they want measured. For example, the 50% quartile for an advert would have a unique URL that is pinged by the player once it reaches -- you guessed it -- 50% through the ad.
Aiding effectiveness is the ability for the ad server to target the video ad to the viewer, including through keywords that may identify the page more precisely, as well as the referring URL and the device.
Figure 2 outlines how basic online video advertising works. What I’ve described is pretty much what Video Ad Serving Template (VAST) defines. The Interactive Advertising Bureau (IAB) -- a coalition of interested parties -- defined VAST to ensure that video players from different vendors could request and receive adverts from different ad networks and play them back consistently. Mr. Spock would approve.
Figure 2. A breakdown of tracking and targeting elements in a typical video advertising environment
As well as pre-rolls, VAST supports mid-rolls and post-rolls, together with companion ads (graphics that may be placed around the video player). VAST also supports an ad network calling out to another ad network -- if an ad network doesn’t have an ad, it can ask a connected ad network, “Do you have an ad I can show?” In case of a string of empty responses, a default failover ad can be programmed in the player (although not strictly provided by VAST).
So, if that’s VAST, what’s VPAID? It stands for Video Player AdServing Interface Definition. It builds on VAST to allow interactive ads in a video player. At the simplest level, consider allowing various regions to be clicked on, for example a pre-roll about a product with a clickable “Learn more” button followed by the main feature with a small “Learn more” graphic superimposed over it -- all played back within the video player.
The online video advertising market has been slow to take off, but with the growth in tablet usage and social media—not to mention real-time bidding and other innovations in advertising networks and formats—it's finally moving toward maturity.
While video viewing on devices is increasing, advertisers aren't reaching the increasingly fragmented audience.