Addressing the problem of accessibility in ever-growing VOD libraries

Video-on-demand has seen ever increasing numbers since its first deployments in 1999 en route to the 250 million monthly views received by Comcast alone in July 2007. However, as the popularity of this service continues to increase, VOD is becoming a victim of its own success, with a growing number of denied requests, increasing support costs for managing aging systems, and rising costs for marketing all the titles within growing libraries.

While upgrading the aging VOD platforms and allocating additional QAMs to the VOD service easily address the first two, the challenge of marketing the far reaches of VOD’s growing library represents a problem that cannot be solved as easily.

Limitations of screen real estate
In VOD, the location of the title is very important in effectively marketing a growing list of video titles. The available screen real estate combined with the lower text resolution in standard definition (SD) television, however, limits the number of titles that can be effectively displayed on any resulting search to four to nine programs, depending on how much additional information is provided (e.g. title, box art, director, etc.).

High definition (HD) TV offers a marginal improvement by allowing for six to 15 titles to be displayed as part of a program search. But the high ranges of these numbers can be misleading due to the fact that listing more titles on a display means less additional information can accompany the program title in the display – thus reducing the effectiveness of each title in the listing.

Figure 1
Figure 1: There is only so much space for data on each menu screen, forcing viewers to drill
down through multiple screens to get more information on the choices they have.
Source: SeaChange International

Besides marketing titles within a dedicated search function of VOD found in the electronic program guide (EPG), operators are experimenting with various ways to slice and dice their library of available titles to make them more accessible (such as categorizing titles into genres or series). Genre and series filters are especially useful for accessing and listing available programs related to currently running broadcast programs displayed within the EPG or program schedule, but they don’t address unrelated long tail content that can quickly become stale in the library due to its lack of recollection by the consumer or association with another currently broadcasted program.

VOD should benefit from SDV
VOD choices are also limited by how much memory individual STBs have to dedicate to VOD listing data, so that the data is readily accessible. The preferred choice is to push this metadata in a compressed format to STBs so that titles can be called up quickly from within the EPG.

However, as the number of VOD titles expand, the memory available must be divided among increasing numbers of titles, restricting the amount of metadata that can be associated with any one title in order for all titles to be stored on the STB.

“As the number of titles (linear or, more pertinent, on-demand) increases, cable server-hosted search scenarios become much more compelling,” says Sean Duggan, senior product manager of Digeo.

These remote search applications are now available from every VOD vendor and provide adequate search capabilities for title libraries too large to push down to STBs – which results in a viable interim solution for VOD operators with greater than 1,000 titles.

As switched digital video (SDV) becomes ubiquitous, a much more feature-rich VOD search interface should evolve at the data center with no more than a 100 ms delay in registering remote control button pushes on the display. With this “hosted” search interface, intelligent searches can be completed, allowing real time searches to take place with each key press. That can include many other optional drop down program filters – by genre, release year, etc.

Such a display should enable VOD operators to expand their libraries to several thousand titles while providing a respectable level of accessibility. However, competing with Netflix’s 80,000 title (and growing) library is currently out of the scope for current VOD services with searches limited to the confines of the television screen.

Using second and third screens
In light of the low text resolution of SD and even HD from a VOD title listing standpoint, VOD will need to look to second and third screens to make its expanding library more accessible.

“VOD movies scheduled via a computer for some time in the future (like current pay-per-view) will provide the high-resolution screen needed to access large VOD libraries,” says David Stengle, VP of Distribution at Black Arrow.

Figure 2
Figure 2: Increasing demand – total U.S. video-on-demand orders since 2004.

It is believed that offering such a second screen as well as a cell phone screen to tap into more spontaneous movie selection requests such as those generated while shopping, or while reading a billboard could be a great way to maintain a “must-watch list.” Such means of movie selections for VOD will provide an effective means of mining the far reaches of a comprehensive video library while further coupling multiple services – but in a different way than currently utilized because such selections happen “off-line” or outside of an engaging VOD session that ends with your viewing the selected movie.

Such “off-line” selections would ultimately enter a Netflix type queue or “my must-watch list” that can be accessed from the living room at the subscriber’s convenience, much like a digital video recorder’s (DVR) list of stored programs. A VOD “must-watch list” could be recalled by the VOD user interface so the consumer could select any title from their “must-watch list” (not all of which may yet be available on VOD) and watch it immediately rather than run to the store or wait for tomorrow’s mail.

In fact, the same queue technology could result in a pretty interesting pseudo VOD offering by satellite companies if they could proactively download queued entries to the customer’s DVR.

Differentiator: Number of titles
With the sheer numbers of titles becoming available in multiple formats (SD and HD), the notion of building out space for every available title begins to raise an interesting business question: At what point does it cost more to buy and maintain the additional disk space for any given title versus what that title will capture in revenue?

Clearly a VOD title wouldn’t be added to the library if there wasn’t some chance that people would rent it, but when it comes to long tail content, would one or two views a year justify the cost of storing both formats of that title? Well, maybe...

One might argue that once a title has been ingested and resources (disk space) have been allocated to store it, the major expense of hosting the title is behind you. However, as disk drives begin to fail and need replacing, current disk arrays need upgrading to faster ones, large video libraries require dedicated 24-hour staffing, etc., the real costs of maintaining large video libraries will become evident such that the revenue opportunity versus cost of hosting each title will be a factor in whether the title stays or gets archived or overwritten.

It is reasonable to assume that beyond the question of whether a title remains in the library or not, additional storage conscious options will be available to VOD operators. For example, VOD operators may elect to only store an SD version of the title rather than both HD and SD versions based on its viewing popularity.

It is realistic that after further business analysis, VOD operators will end up focusing on a very limited window of available titles (see Figure 1). VOD will likely end up covering all new releases and ride on the coattails of the marketing blitz that studios run as part of releasing new titles for purchase. These titles, along with subscription VOD and other content deals each VOD operator makes with their content providers, will make up a majority of its VOD library. Those titles that maintain some reasonable rental frequency, say 5-10 rentals a month, would likely remain within their video library due to popular demand, whereas those titles viewed much less frequently would either get pushed to some archive or get overwritten.

To effectively market their growing library, VOD operators must exploit the best features of online movie rentals, providing the ultimate in convenience, all the while zeroing in on movie rental’s Achilles heel – time delays in obtaining the next title from the queue. While it may not ever be realistic for VOD to command an 80,000-title library, a reasonably deep video library with instant access must-watch lists; accessed through second and third screens; opt-out windows that would allow viewers to change their mind about a movie within the first five to 10 minutes without paying for it; a high-quality viewing experience; and instant access to the desired title with simplicity and not all the “forced previews” of DVDs would provide a formidable service that would not only compete with best-in-class online rental companies but beat them with ultimate convenience.

Such a service could relegate the likes of Netflix and Blockbuster to mere niche players who provide deep libraries of hard-to-find titles. While this too could also be a lucrative business for some time, VOD operators need storage technology to get much better and significantly cheaper for them to cost-effectively build and maintain a 100,000-plus title library.