The Economics of Live Events

The advent of the 21st century's second decade offers the opportunity to look back to gauge the newsmakers and news breakers of 2000-2009.

In the streaming world, where we're now into the 14th year, the pace of innovation means that new technologies are being tried, and old technologies are being retested-some for the third or fourth time, depending on the corporate and individual memories of these technologies and on previous business execution attempts.

One area that has remained unchallenged, though, is the acceptance of large-scale live event streaming. Streaming, after all, began as a way to deliver audio across the internet (or an intranet) in real time to those who wanted to hear about a particular, timely topic. From concerts to breaking news to sports, streaming's roots are in live broadcasts.

Fast-forward to 2010, though, and we're in an almost completely different delivery model. CDNs routinely state that 95% of what they deliver is on-demand content. Even if the on-demand content is streamed-using some of the newer HTTP streaming and adaptive bitrate solutions-it is still prerecorded content that has been transcoded and uploaded as a file to a server for delivery.

As I noted in the Streaming Media article "Navigating the Enterprise Workflow" (October/November 2009), many corporations are moving toward segmenting content from all-hands meetings or executive town hall meetings into topical bite-size chunks, as a way to get additional value out of content that used to only be shown during the live broadcast and subsequent rebroadcasts.

Beyond the corporation, though, we're also past the point where it was difficult to make the technology work for large audiences of thousands or even millions. Now, events are routinely seen by audiences ranging from 4 million-9 million, although often accompanied by some level of frustration on the part of a small number of viewers (or a large number in rare instances).

From last year's presidential inauguration to Oprah Winfrey's 2009 webcasts to the Michael Jackson memorial service, the web seems to be fairly resilient in being able to absorb viewership numbers similar to what you'd see at the low end of the Nielsen television ratings.

So why not leave well enough alone and avoid the "third rail" discussion of the live streaming of large events? As I mentioned at the outset, it's beneficial to look back at where we started to make sure our assumptions are still valid.
In addition, with the emergence of true HD streaming for large-scale events such as the Winter Olympics, it's important to see whether we're complementing or competing against traditional broadcast, which is also moving to HD delivery.
Regardless of streaming's roots as a technology to deliver events as they happen, large-scale live webcasts aren't cheap. From the production costs to the cost of delivering the bits to thousands or millions of concurrent viewers or listeners, live events simply cost more than their on-demand counterparts do. And then there are the intangible costs: If all goes well, everybody's happy. But if there are problems, the bad publicity resulting from a poor user experience besmirches not only the parties involved in the particular event but the entire industry.

For this article, I did an informal survey of some of the people and companies involved in both sides of the live event streaming equation-production and delivery. By interviewing those with insight into how the other half of the equation operates, I hoped to gauge the commonality of the perceived areas of opportunity and optimization.

While the sample set was limited-and the survey unscientific-I was surprised by the amount of information received as well as the overwhelming interest in the topic. I'd like to revisit this topic with a more formal survey at some point, with results generated as a white paper or as an additional set of articles.

I also received a great deal of interest from the ecosystem companies, such as hardware encoders and DRM providers-enough interest, in fact, that we'll explore this middle ground in a future article that will deal with the ROI of a full-spectrum case study event.

Market Segments
I asked the survey respondents to list the market segments in which they found the biggest traction for live events. I suggested six market segments (news, government, corporate, concerts, pay per view, and sports), but I encouraged them to add other key market segments.

Tallying up all the responses, sports and news dominated the top end of the list, both on the delivery side and on the production side. An aggregated ranking might look like the following list, with the percentages representing the overall ranking across the multiple respondents:

• News: Breaking (90%)
• Sports (75%)
• Concerts (75%)
• Corporate (50%)
• Pay per view (50%)
• Government (25%)
• News: Press conferences (25%)

Among those companies that deliver the content bits and can most accurately gauge audience sizes, the most popular types of events are often based on the particular focus of the CDN.

"Sports is by far the biggest segment for events," said Mark Taylor, VP of product for the content markets division at Level 3. "Second is news or one-off, nonsporting events like presidential debates or Michael Jackson's funeral."

"News clearly leads the pack," said Matt Smith, who spent several years at Yahoo! Broadcast Solutions before joining Inlet Technologies as senior director of systems architecture. "When an event is newsworthy and breaking, users have clearly shown they will turn to the digital domain for live, immediate updates and video. Sports is next, as events like March Madness and the U.S. Open have shown, office-bound viewers will flock online to catch a few (or even several) minutes of sports."

Both Yahoo! and Level 3 have broadcast capabilities beyond just the streaming portion, but they often use production companies to augment for particular events. Yet even the production companies have different opinions about the current market segment sweet spot.

One of the production companies, Net-StreamLive, sees news, sports, and concerts as the highest-ranking market segments but also included corporate events such as town hall meetings or all-hands meetings that were dominant in their customer base.

By contrast, Level 3 had an interesting take on corporate events.

"We no longer make a concerted effort to support events in the corporate environment, as the majority of the cost is associated with the production," said Taylor, adding that, compared to sporting events, "the audience is usually small. As such, the corporate segment does not currently match our capabilities."

This is not to say that Level 3 does not see value in the corporate space, as Taylor and Peter Neill, senior VP of content markets, also noted that they "do a number of these events every month within Level 3" for internal consumption.
Yet their point is that corporate all-hands meetings might not be classified as large-scale events, if the live stream only ends up being seen by 5,000-15,000 simultaneous viewers.

Even in a company that has 100,000 or more employees, the lower number of simultaneous viewers can be explained, somewhat, by shifts in viewing habits that have occurred over the past few years. In corporations that have had streaming in place for more than a year, many employees have a choice as to where and when they can watch an event-in real time or at a later date as on-demand content-and very few corporations provide scheduled rebroadcasting, which was a staple of the early days of corporate webcasts.

Rebroadcasting was itself a vestige from televised broadcasts in the days before TiVo: A webcast stream would be captured during the live event and then streamed again once or twice in the days that followed; those who missed a webcast could then see it in its entirety.

What about pay-per-view entertainment and sports events? By and large, the majority of live delivery still operates under what could be referred to as "free to air" or ad-driven models. Yet several respondents said that pay-per-view models are starting to expand into sizable audiences, contingent on the consumer's cost, type of content, and exclusivity of content.

"Within the sports segment, the vast majority of the events we stream are free to air rather than pay per view," said Taylor. "We are starting to see a combination of the two, however, with a lower-quality free stream (or limited, free content) plus a paid, higher-quality stream or more content."