The amount of pre-roll video advertising inventory available for real-time bidding grew by 14% per month in Q2, topping 29.9m streams per day in July according to data from TubeMogul.
As a result of the increase in ad space CPMs declined to £4.97 in Q2 from £6.47 in Q1.
The growth of the RTB video market is a trend identified in our new Real-Time Bidding Buyer’s Guide. In the US alone, RTB will account for around 15% of online video spend in 2012 and is forecast to account for 22% in 2013.
The increase in spending is partly attributed to the fact that the capabilities around audience buying and creative optimisation are a natural fit for video advertising.
TubeMogul’s Quarterly Research Report also shows that CPMs don’t match up with viewing patterns, with the biggest divergence occurring over the weekend and early in the week.
Weekends represent 29.4% of the week’s video viewing but have the lowest prevailing CPMs at £5.40, so there is potentially an opportunity for marketers to run more cost-effective video campaigns.
The report also looks at the completion and click-through rates achieved by 30 second ads on different types of websites.
TubeMogul ranks sites into different tiers based on quality and context. Tier one sites are major broadcasters, officially-streamed music videos, cable networks, sports broadcasts and major news outlets.
Tier two sites are publishers that have high quality content (often high-trafficked online-only outlets), and tier three sites are niche video outlets that often cater to narrower interests.
While completion rates remain fairly consistent across the tiers, CTRs vary widely with tier one sites proving to be the worst performers.
However the report concludes that the variations between site categories are still relatively small, so decisions on which ads to bid on should instead be based on what type of consumers marketers want to target.
The data comes from campaigns run through TubeMogul’s media buying platform, which includes ten of millions of streams from standard pre-roll ads.