Myth #3: Online video is easy to track, evaluate and optimize.
Reality: While the IAB has made great strides in determining online video ad guidelines, there are still no standards for format, delivery or reporting. While many video ad units are clickable, there is also a segment that is not, particularly for pre-roll ads. As a result, advertisers need to deliver adjacent ads that provide an opportunity for interaction, typically a flash or GIF unit. Often, these two ads, though part of the same campaign, are delivered by different servers -- one for the video ads and one for the adjacent units. While DoubleClick and other companies claim to offer centralized serving products for video, the technology has not quite caught up to the need.

So, while an advertiser can create a single video ad unit to run across multiple vendors, there is still no standardized reporting. Even the most rudimentary tracking and reporting is extremely difficult -- and central serving, rather than simplifying the process, exacerbates the problem due to very wide discrepancies in reporting impressions. Thus, while inventory is sold and video advertising is hyped, the reporting technology and trackability lags behind advertiser adoption.

When an advertiser uses multiple video vendors, the problems are compounded since each uses a different reporting platform. The fact that some offer clickable pre-roll units while others don't makes it extremely difficult to create benchmarks for performance across the entire campaign. Agencies complain that providing a basic campaign post-mortem requires too much work and, given the manpower required, is difficult to justify. 

Myth #4: Online video advertising should use the same branding metrics used for television.
Reality: Traditional branding metrics only tell a small piece of the story. Online branding is not just a measure of awareness and affinity. Advertisers are blessed (and quite frankly, cursed) with the ability to track a great deal more in their units than impressions and clicks. Marketers should use engagement and ad interaction to not just evaluate direct response, but also measure branding. Site-side behavior post-impression should also be used as a measure. Site-side behavior can be used to deduce branding -- metrics such as usage intensity of visit, repeat sessions, page views, frequency and time spent. These should be used to quantify the brand impact of an ad.