Targeting relevant ads through streaming media (video) has been a focal point of innovation for many startups over the past few years. There is good reason for this. If ads are made more relevant to consumers conversion rates will increase, driving down acquisition costs. Lower acquisition costs attract more advertising dollars and increase willingness to pay for CPMs. This in turn makes the auction for advertising space more intense, driving up prices and making both publishers and advertiser networks wealthier. The bottom-line is that better targeting will increase revenues for Internet companies.
To date, the methods of targeting ads through streaming media remains less effective than the methods used for textual advertising. This is because video is much more difficult for a computer to interpret than text. Computers can't currently decipher the conent video clips on a large scale (millions of files). As a result, several work-arounds had been created. One such strategy was to target video ads based upon the focus of the webpage. For example, ads about cat food might be shown on a pet site. But do dog-lovers that are watching a video about a dog want to hear about cat food? Further, what happens when Purina's ad shows up before some inappropriate video that a user uploaded as a prank?
However before you lose too much sleep, you should know that there appear to be some solutions in the making. The WSJ reported last week in an article titled Surfing the Internet for Spoken Words (The WSJ link doesn't work w/o a subscription) that several companies are developing technologies that will use speech recognition to evaluate video content and determine if it is both appropriate and relevant for an advertisement. If this technology is successful it could serve as a tremendous competitive advantage to advertisers networks.
If it works well, I would assume that these companies will be purchased by some of the big boys. Will YouTube be one such buyer? They could become the google for video - the largest publisher and ad network in one...
No comments:
Post a Comment