Cracking the Online Video Monetization Nut: Adap.tv CEO Amir Ashkenazi (Part 7)
Check other articles in the series...We continue our discussion about the technical and strategic implementation of the Adapt.tv technology. The core behind it is a learning system which is fed data from contextual matching as well as inputs from user interaction with ads. Over time, this produces an optimized ad campaign for video-content providers, addressing a validated market need.
SM: How will you be able to determine if the ad you push is the most effective? AA: There are two parts to the magic. One is the contextual matching, and the other is the learning system. We monitor user interaction with the ads very closely. We use the information to optimize the ad selection. Ads which are not effective will be eliminated, and ads which are effective will be displayed more frequently.
SM: There are multiple questions that brings to mind. The first is what is the measure of effectiveness? Second, video is much more contextually complex to analyze than text search, how can you do that? AA: Absolutely, and this is why we have some real rocket scientist on the team. Ben Klein is a well known researcher in the area of machine learning and natural language processing and is a professor at Berkley.
SM: Those are technical solutions, and to me those are implementation details. What I am trying to get to is the philosophy of your behavioral targeting or contextual targeting. What is the underlying philosophy? AA: The effectiveness is measured by yield. By revenue generated.
SM: Is yield a function of click through or is there some other function? AA: We support different kinds of ads including CPC, CPA and CPM. Those are all valid pricing models we can work with. The algorithm is smart enough to use all of those and optimize the ad selection to maximize the overall revenue generated per minute of video. At the end of the day our algorithms are being measured by revenue generation.
SM: Sorry to push you, but what is the philosophy behind the targeting? AA: It is related to how we maximize effectiveness. At the end of the day advertisers want to maximize the effectiveness of their campaign, publishers want to maximize revenue, and viewers want to have non-intrusive and relevant ads displayed. We keep all three in mind when selecting the right ads at the right time.
This segment is part 7 in a 10 part series Segment 8 →
Jump to part: 1, 2, 3, 4, 5, 6, 7, 8, 9, 10
[...] [Part 7] [Part 6] [Part 5] [Part 4] [Part 3] [Part 2] [Part 1] [...]
[...] 8] [Part 7] [Part 6] [Part 5] [Part 4] [Part 3] [Part 2] [Part [...]