As the broadband video world continues to coalesce around advertising as its primary business model, there is a flurry of companies seeking to improve the monetization process. As I've written before, this is critical work, because at some point the bloom will be off the broadband video rose if participants can't earn an attractive ROI.
CEO/co-founder Amir Ashkenazi recently gave me a run-down on Adap.tv's approach and progress. Amir was the founder of Shopping.com, which was acquired by eBay and he has brought together many former colleagues for his experienced management team.
Like its competitors, the heart of Adap.tv's model is its ad targeting and relevance engine. Adap.tv uses a "multi-disciplinary approach": analysis of the video/audio (context, metadata, etc.), analysis of the ad (keyword submission, etc.) and analysis of the user (demographics, location, etc.). This data is then fed to a matching engine to pair ads with the most relevant video. Over time the system optimizes based on actual click behavior.
Adap.tv is highly focused on overlays (Amir believes this will be the "de-facto standard" soon), and provides a series of customizable templates for advertisers (see below Kayak overlay). It is also positioning itself as a cost-per-click model, so there's no fixed cost to advertisers. In fact, advertisers can power Adap.tv ads using the same keyword feeds they use for their keyword campaigns.
So far publishers have been responsive to the CPC model because they see overlays as opening up a lot of untapped inventory. Obviously implementing overlays needs to be done judiciously or the viewer experience will become cluttered and broken. Amir believes the whole broadband video ad model will move to CPC over time as advertisers become more sophisticated and focused on performance. This Google-like model would be very good news for advertisers, but would be a brave new world for traditional broadcast and cable networks long accustomed to CPM approaches in their traditional businesses.
While I think a more performance-based broadband ad environment would be welcome, I continue to believe a CPC/overlay approaches will ultimately co-exist with CPM/pre-rolls. There's a lot of interest in overlays, yet there are too many great 15 and 30 second TV spots not be re-used online and the CPMs are way too rich for big branded content providers to walk away from.
Other companies that are in the contextual analysis and/or overlay space include: ScanScout, Digitalsmiths (note: a VideoNuze sponsor), YuMe, blinkx, VideoEgg, YouTube, Brightcove, AdBrite, Viddler (which TechCrunch just wrote about yesterday) and others I'm sure I'm missing or are yet to surface.
One of the things I really enjoy about being an analyst in the burgeoning broadband video industry is getting first-hand exposure to all the clever innovation that's going on. I find it endlessly fascinating to hear directly from entrepreneurs on the front lines where the kernel of their idea came from which led to their business plan. A user experience issue? A technology deficiency? A business model flaw? Over the years I've heard many stories. Some kernels have real weight, while some don't quite resonate for me.
Magnify.net falls into the former category. My read is that this is a company trying to solve a real problem with a very clever solution and the right "corporate attitude" to make it a likely winner.
Magnify is actually solving a number of real problems, many of which relate to the highly distributed or "Long Tail" nature of the Internet and broadband video. First is that while consumers love broadband video, finding what they want is problematic. Novelty quickly turns to frustration when rummaging through big video sharing sites to find something relevant. No matter how much users want choice, some level of editorial or "curation" is essential to optimize their experience.
Magnify enables existing enthusiast or vertical web sites (whether independent or major media) to obtain video from the best video sharing sites (YouTube, Metacafe, etc.) and coherently present a screened assortment to their users. The sites' use their editorial skills to sort the wheat from the chafe, with easy-to-use admin tools ensuring that no offending video slips through the cracks.
So the second problem Magnify solves is enabling thousands (17,500 and counting to be exact) of sites to provide quality video to their users without the hassle and expense of creating it themselves (the "matchmaker" role). These sites get 50% of the revenue from the ads Magnify sells around the video (or they can keep up to 50% of the inventory to sell themselves), leveraging their audience and subject matter expertise. Incorporating video into web sites is becoming online table stakes. I agree with Steve, in the years ahead, sites without video are going to look "charming".
The only real hole I can find in Magnify's model is that it doesn't currently compensate the content creators themselves (a la Revver for example). However I'd expect that to change as creators upload directly to Magnify and the company's network and traffic builds out over time.
Lastly, I like Steve's attitude. He views the market as an incredibly expanding pie, and not "winner take all." As a result, while there are others who touch on Magnify's space (Brightcove, ROO, VideoEgg, Ning, KickApps, etc.), he's less concerned about competition per se and matching feature-for-feature, but rather on responding to the needs and wants expressed directly by their own user base. Companies that do this ultimately win, regardless of competition.
The Magnify story plays into a number of areas I follow closely - the changing role and power of video distributors, the continued "nichification" of video, the challenge of video discovery and the reliance on ads, not subscription fees. To the extent that their approach succeeds it will further morph traditional video models. For a 10 person company that's only done an angel round, they've accomplished a lot in addressing genuine Long Tail issues in the broadband video industry. (Btw, TechCrunch has 2 great reviews, here and here).