What YouTube started has now spread all over the Web and likely across your network. A tablet with iMovie or Avid Studio can turn anyone into a TV producer--in fact, a typical salesperson with an iPhone and iPad arguably has more video-capture and production technology muscle at his fingertips than TV studios did just 20 years ago.
For roughly 15% of respondents to our 2012 InformationWeek State of Storage Survey, rich media (broadly defined as video, audio, and images, of which video is by far the largest file type) is gobbling Tier 2 and Tier 3 storage space at a good clip. And video and imaging data isn't just big, it's persistent: About a third of respondents save rich media content for two to eight years, 20% keep these files indefinitely, and 30% don't have a policy--which usually translates to "forever."
Video, while undeniably popular, has some nasty attributes: The files aren't easily searchable, and the content tends to originate from a hodgepodge of special-purpose applications that are often procured and operated under our radar. Ballooning video stores could mean a return to the data silos we've spent a lot of time and energy busting up, something no one wants. The only way to avoid that is to manage video as a data asset; after all, it's potentially a treasure trove of idle knowledge.
Surveillance systems are probably the largest source of video data. Our 2012 InformationWeek Physical/Logical Security Survey shows that cameras are second only to fire and burglar alarms among physical security systems respondents have in place. The importance of this data is clear: 61% have reviewed video security footage or provided it to law enforcement.
Making The Case
While storage has gotten exponentially less expensive, it's far from free, and compared with a typical Office document or PDF, which might weigh in at 100 KB or so, video files are gargantuan. A standard-definition YouTube video runs about 1 Mbps, with 1080p high definition coming in at 3 to 4 Mbps, so a five-minute SD clip uses about 35 to 40 MB, while an HD version takes 100 to 150 MB.
These sizes, high growth rates, and lengthy retention periods give CIOs ammunition to get a plan in place. To do that, consider two main concerns: How and where do we store larger and larger video data sets? That's an infrastructure problem. And how can we help the business leverage video data by making it sharable, searchable, and reusable? That's an information and content management problem.
To solve the first challenge means building a storage architecture with systems optimized for video that can be easily, rapidly, and incrementally expanded using low-cost hardware. The second requires software that can make a video clip as easy to find as an archived email.
The first step is to look at which applications and departments are generating video data. The usual suspects include video camera image capture and editing (for things like marketing and training), IP surveillance, and videoconferencing. Of these, the last is the most likely to be under IT's direct purview, but it also generates the smallest amount of stored data since videoconferences are often transitory.
Still, there's growth looming in most areas. Our 2012 InformationWeek Consumerization of IT Survey found that nearly half of 400 respondents have room-based videoconferencing in production, and 32% are planning or piloting it. With single-user desktop videoconferencing, 70% are in planning, pilot, or production.