Forrester's Josh Bernoff thinks video search stinks -- and he's right. With all of the focus on putting video content online and growing viewership, why does it seem like so few companies are focused on fixing a major usability problem that could hobble the digital video industry?
Although some may see Google's YouTube acquisition as a sure sign of video's point-of-arrival, we are merely in the early stages of the digital video revolution and we have a long way to go. Note that I say "digital video" rather than "online video" -- online video will continue to evolve and grow over the coming years, but we are also in the early stages of a long series of disruptive changes to the way video is delivered and consumed across all possible screens.
What's happening in video now?
The long tail gets longer by the day -- note today's launch of Travelistic, a social video site focused specifically on travel for the 25-34 set and only the first in a series of online video plays to be launched by MTV veteran Nicholas Butterworth's Diversion Media. There are literally hundreds of video sharing destinations and we are now moving beyond general interest aggregators to sites targeting not only specific demographics but specific lifestyles, passions and affinities.
Mainstream media companies are accelerating their online video initiatives -- they're repurposing content from their offline shows, posting outtakes and exclusive extras, and in some cases creating web-only programming to beef up their online presence. A new wave of start-ups are forging partnerships with the mainstream media companies to develop new video delivery and syndication models. (See this week's BusinessWeek article on this topic - it focuses on The Venice Project, a broadband long-form video play from the founders of Kazaa and Skype.)
The big web portals are clueless about video -- Forrester's Bernoff argues that, in their misguided attempts to be all things to all people and their equally misguided attempts to ape YouTube's success in user generated video, the major portals have failed to create meaningful differentiation in their video offerings and more importantly have build video services that are all but unusable. The illustration below highlights the confused (and confusing) and confusion state of affairs. Is AOL a destination for premium video streams, paid downloads, video search or the sharing of user submitted videos? YES!
So with the universe of online video getting bigger and more complicated every day, how does a consumer go about finding the content they might actually be interested in watching? The short answer? They don't.
The longer answer? Here's an extract from Forrester's report illustrating exactly how and why video search fails consumers:
"Suppose you were seeking the trailer of the recently released cult film "A Scanner Darkly." You might find it — or you might get directed to a broken link or to low-quality footage of an interview with the late writer Philip K. Dick from 1974 (because the movie is based on his book). Looking for the free video preview of NBC's "Studio 60 on the Sunset Strip"? Video search engines can't find it. Instead, they will route you to cast interviews and, bizarrely, an interview with Ashlee Simpson that has nothing at all to do with the show (see illustration below). These frustrating results are typical; if you're looking for high-quality video on the Internet, video search is hit-or-miss. Why such poor results? Because:
It may not be clear what the searcher wants. If you search for a movie title, do you want the music video, a clip, the trailer, or the interviews with the actors on "Entertainment Tonight?" Video search engines can't tell, so they guess and — lacking any other indicators of user intent (or clairvoyance) — guess wrong.
Video search rankings can't leverage links for relevancy. A user-generated parody posted on YouTube may get ranked just as high as an actual trailer, especially if more people link to the parody. The most-linked sites float to the top of Google rankings, and appropriately so for the most part. But seekers for Internet video may not be looking for the most-linked video.
Search engines favor their own content. AOL's top three results were on AOL sites, while Lycos' blinkx-driven searches favored sites like ITN and IFILM with which blinkx has relationships. Meanwhile, all the search engines missed the free video on NBC.com, the actual target of our "Studio 60" search. If search engines mostly find their own content, consumers won't trust them.
The same video has endless duplicates online. Professional video producers will often syndicate material to multiple sites — why not get more exposure for promotional or ad-supported content? The result: search results that point to duplicate material that gunk up the search index. This same challenging problem leads users — in the case of Yahoo!'s video search — to user sites that can't handle the traffic.
Metadata about video is inconsistent. Unlike text, video content doesn't include words to index. Instead, search engines are forced to look at the words that surround the video or information posted in Media RSS or other standardized metadata formats that tag along with the video. The problem: Video creators don't do a consistent job in providing adequate and accurate metadata. Some search engines, such as PODZINGER and blinkx, actually index the video's audio track, using voice recognition and looking for sounds that match recognizable words. But even blinkx, which combines all of these techniques, got fooled and took us to a trailer for the wrong movie in our "A Scanner Darkly" search."
BOTTOM LINE: In the million channel universe, how do you actually find the one channel (or even the one program on the one channel) you care about? Somebody needs to figure this out quick, or digital video could collapse under it's own weight.
