Industry Perspectives: The Promise of Open Source Video

[Editor's Note: Industry Perspectives is a regular feature in which vendors in the streaming media space explore issues and trends on which they can shed unique perspective. The articles reflect the opinions of the authors only, and we print them as a means of starting discussion.]

As we enter 2009 three things about online video become clear: (1) with over 17 Billion videos watched in December 2008 in the U.S. alone (according to comScore) and steady month-over-month growth, video is here to stay; (2) video exhibits classic long-tail distribution—while YouTube remains the dominant player, video is rapidly moving from destination sites to the rest of the web, with millions of sites streaming video as the new mode of communication; and (3) the conversation at all levels is shifting from the technological aspects to the value aspects: not how to build a player or convert between formats but, rather, how to foster audience engagement and monetize on these billions of streams.

These observations are underwritten by the trends that became apparent in the past several years, namely, the rapid reduction in the cost of delivery and the abundance of companies in this ecosystem. For anyone who is part of the video universe, the key question that remains open is what drives value in this brave new world. How can publishers, advertisers, and technology enablers make money in a world in which delivery (CDN) is commoditized, display opportunities are abundant (driving CPMs for video advertising down), and audiences expect to get everything for free? The short answer, I believe, is to focus on innovation—of formats, user experience, content, or delivery.

And here is where open source video enters the picture: It is a development methodology and distribution strategy that allows each company in the ecosystem to focus on what it does best, instead of replicating the efforts of others. Open source video, introduced to the market by companies such as Kaltura, is being adopted at every level of the ecosystem by industry leaders such as Akamai, Mozilla, and Wikipedia. Its premise is simple: Video is too important of a medium to be controlled by a single player. By espousing the principles of openness at all levels, including formats, technology, and content, and by collaborating in the development process, video can enjoy the force-multipliers that we have seen in other areas of open source software. The result is a better user experience, a reduction in the total cost of ownership, and a focus on innovative value-driven results.

To understand these last claims, let us unpack the essential elements in the video solution stack. Those are divided into several inter-related areas of functionality:

Ingestion & Transcoding
Every video solution begins with the ingestion of new content. Every solution needs the ability to bring in new content, sometimes from a variety of sources (like uploading files or recording from a webcam), and to convert them into formats that are appropriate for streaming and playback. The need to prepare files for later download and offline use, or for mobile delivery, adds an extra layer of complexity. An open source solution guarantees maximum support for file types, multiple transcoding options into a variety of formats, and the prevention of lock-in effects into proprietary formats.

Content Management
In this layer, the solution needs to handle the content and the metadata around the content. This includes quality control, moderation, flagging of problematic content, channel building, playlist building, taxonomy, folksonomy, metadata extraction, and anything else that insures that the right content is parsed properly and is ready for playback. An open approach for video management focuses on the ability to integrate easily with other CMS systems, and includes the ability to synchronize meta-data with host systems, so that video results are optimized for search engines (SEO) and appear to be integrated with the rest of the site.

Hosting & Streaming
As we know, video content is huge in size and over time even a simple video solution will face challenges of handling large volumes of large files. This layer in the stack needs to handle this problem in a cost-efficient way, meaning that files need to be stored in the right formats, disseminated over the content delivery network in an efficient manner, and be streamed in a smart way that understands the user environment and needs. Privacy and security concerns add an additional layer of complexity. An open approach to streaming and hosting will allow publishers to choose whether they want to self-host, or enjoy the ease of use of a Software as a Service provider. It should also allow publishers to migrate their content if they change their decision, and allow flexibility in the choice of content delivery networks.

Player Development, Publishing, and Syndication
The three layers of this stack take care of everything "behind the scenes" but, eventually, video streaming is a user-focused activity. All the preparations of the content would count for nothing without the tools and services that publish the content. The core element in this layer is the video player itself, which can be branded and which is usually integrated within a larger user experience that allows users not only to watch the video but also to interact with it by adding comments, voting, rating, tagging, and more. An essential element in this experience is the ability to share and syndicate the content, either by the publisher or by the end user. The syndication options depend in large part on licensing rights and choices about the level of control that authors want over their content. An open source solution to this layer of the stack should allow publishers to easily develop custom players, to gain access to large repositories of content with appropriate license information (for free or for a fee), and to be able to easily control the distribution and syndication of their content across the network.

Reports and Analytics
Websites live and die by their ability to engage their audiences, and the first part of engaging the audience is to understand its composition. Who is watching the video? Where are they coming from? What are the viewing patterns? What correlations exist between demographics and content? All those questions are essential for any media-focused site. This layer of the stack provides tools to answer some of these questions. It includes client-side components that collect the data, and server-side components that aggregate it and analyze it. An open source version should allow all of the above to be monitored centrally, as well as to connect to existing analytics system so that publishers can use the same analytics system for their video and non-video needs.

Monetization
Perhaps the most important layer in the stack for many of the stakeholders is the monetization piece. There are only two business models in media: You either sell content to viewers (pay-per-view, pay-per-download, DVD, subscription) or you sell the audience to advertisers (sponsorships, direct response, brand-advertising, product placement, video search). There is no third model. Any video solution needs to enable either of these or both. An open source monetization layer should be open enough to accommodate all leading ad engines, and, especially in an area of rapid innovation, to be future-proof, allowing new monetization models to be developed on the fly.

So we see that a fully featured video stack is a very complex matter, which is unlikely to be achieved by a single company in one shot. Early proprietary attempts in the market have resulted in heavy and convoluted designs that are expensive to set up and operate. Moreover, these solutions are tailored to a very small group of large media companies who can foot the hefty bills. The rest of the market is struggling to find cost-effective solutions that can meet complex requirements for reasonable costs. The commercial-off-the-shelf (COTS) platforms just don’t cut it. They usually are missing specific functionality that is essential for a seamless integration, and any attempt to tamper with them is expensive and lengthy. As a result, when faced with a build-vs-buy decision, medium-sized media companies, publishers of all sorts, and enterprises are often forced to try and tailor custom made solutions and do the integration in-house. This gives them better control but, of course, is more costly, not future-proof, and requires dealing with a multitude of vendors.

Open source video offers an alternative. By creating a global community of developers—both individuals and corporations—who each focus on their own layer of the stack, and by then releasing all the code for free, open source video promises for the first time the previously impossible: a robust infrastructure that is at one and the same time easy to adopt, adapt, and modify, and cheap to deploy and operate. Developers enjoy full flexibility and an open framework to innovate and customize their own solutions while leveraging the community’s work, and publishers benefit from economies of scale—low CDN costs, a network of syndicated content, value-added services and more.

There is a ton of activity around open source video today. Kaltura, a market-leading pioneer in the area of open source video, has partnered with the Participatory Culture Foundation (creators of the popular Miro software), with Wikipedia, with the Mozilla Foundation, as well as 20 other organizations and instigated the Open Video Alliance (OVA), an umbrella organization for stake-holders who share the vision and believe in the future of open source video. The OVA is centered on raising awareness and developing standards that promote open source video and coordinate members’ activities. Other initiatives in the market include Akamai’s as well as several other open source video player and transcoder projects.

Open source video is not only a future promise: it is becoming a value driver in the present. My experience at Kaltura shows that open source video works, today. We help more than 20,000 sites, companies and organizations the world over (including Wikipedia, the UN, Coca Cola, Pepsi, PBS, Universal Studios, and Lionsgate Entertainment to name just a few) to discover the power of the open video solution stack. Whether self-hosted or managed on the cloud, these organizations pay less, get more, remain in control, and hit their market faster. The bottom line: If you are a publisher, a developer, an agency/creative shop, a content owner or a media company, you should join the open source video bandwagon, and start driving more value today.