SharePoint Intranet for NBC Olympics: How We Did It, Part 1

Greetings and aloha! I’ve returned home to glorious Maui, my home, my dog, and my amazing friends and “ohana” (family) after two months on the road and an extraordinary six weeks in London. Now that the Games of the XXX Summer Olympics are over, I can begin to reflect on what we did, how we did it, and lessons learned for enterprises everywhere.

This week, I’d like to start that process of sharing by laying out the requirements we faced and some of the commercial, off-the-shelf tools we used to support our efforts.

Before I do that, let me give a shout-out to the incredible team—numbering close to 4,000—that brings the Olympics to the US audience. Last week I shared some of the statistics related to NBC Olympics ratings and digital media as they stood at that time. Now, with the games concluded, we can say that we were extraordinarily successful.

More than 219 million unique individuals watched the 2012 London Olympics, making it the most-watched event in television history. On a per-night, average-viewer ratings level, it was the most watched non-US Summer Olympics in 36 years, since Montreal, which I actually remember (barely) and was a bit of a different scenario because there were, well, only three channels one could even choose to watch back then!

The digital and social numbers are also mind blowing. Nearly two billion page views and 160 million video streams. And more social activity (comments, posts, tweets, and fans) than the 2012 Super Bowl, the Grammys, Oscars, Golden Globes and all seven games of the 2011 World Series combined.

If you’re a numbers geek or are interested in social or digital media, you should dig into the press releases at NBC Sports. Staggering stuff.

And as I posited last week, despite time zone differences and tape delays, people watched events live on cable and online, and people still sat down to watch the prime time coverage, with its focus on athletes, stories, sports, and quality programming that managed to cram hundreds of hours of competition each day into three or four hours.

I really love the fact that we as a nation can still come together around something like the Olympics. That is what it is all about.

With any project, including a technology implementation like SharePoint, you have to begin with the “requirements”—a word I don’t like to use, but it serves the purpose. You have to understand the desired outcomes (what needs to be accomplished), the constraints (time, money), the resources (people, expertise, tools), and the political realities.

So before I share some of the specific ways we built our environment and what we learned this time around, let’s spend this week laying the foundation of those requirements.

The broadcast of the Olympics on NBC is a herculean effort. As I mentioned, close to 4,000 people are directly involved at NBC with bringing the Games to the US audience, and tens of thousands more bring the games to audiences around the world through broadcast, print, radio, and other media channels. I play a very very small role in that effort. My “credit” for the broadcast is as the “Microsoft Technologies Consultant.” ( The credit ran at 3:57pm on Friday, at the end of the Daytime Show. Did anyone actually see it? Not if you blinked.)
 

What that means is that I help design, implement, and support some of the key Microsoft technologies, including Windows (server and client), identity (Active Directory), management (Group Policy, software deployment, patching), several specific applications, and of course, SharePoint. Preceding and during the Games, I’m able to lend my expertise to the talented, lean team (numbering around five) of IT staff that work full time, year round for NBC Olympics.

Most of these thousands of people come together for a very short time—two weeks or a month or two at most. They come from all parts of NBC, GE, and Comcast.

They come from vendors, like Avid and Sony and Ericsson and Cisco. And they come into an enterprise that has a real life cycle of approximately six weeks.

When I arrived on July 4, there were still walls being built and painted. Twenty days later, we had 3,500 people at work.

Even before Closing Ceremonies—starting Friday—entire teams were packed up and on their way out the door. Sunday night during Closing, rooms were being torn down. And during this time we’re generating insane amounts of content, viewership, and revenue.

You can see a few pictures of the birth and death of our enterprise on my Facebook photo album.

Planning for this effort takes years. We’re already drawing floorplans for Sochi, Russia (Winter Olympics in February, 2014) and looking toward Rio (2016 Summer Olympics). But there is no way to adequately plan or test for what we do.

There are many pieces of the puzzle that come together only at Games time. And, you would be surprised just how much of what we do and what we bring to air has never, ever been done before. So a lot of what we do is invented, on the spot, to solve problems.

Some of our vendors and staff call the Olympics “The Billion Dollar Lab.” That’s a pretty fair statement: It’s a very challenging and dynamic environment. The most critical requirements stem from that reality:

• First, whatever we do must be done quickly. It must work, period. There’s very little time for troubleshooting failure, and virtually zero tolerance for failure of any production system once the Games begin.
• What we build must be easy for end users. There’s no time for training 3500 people who arrive a few days before Games begin. User adoption isn’t really a challenge, because whatever tools we provide must be used, but it must be easy to use because there’s no time to train or troubleshoot end-user problems.
• BYOD must be considered. A related reality that was much more significant in this Games than previous games is the “Bring Your Own Device” syndrome. It hit us big time. We knew it would happen, but it was a bigger challenge than I expected, and in ways I didn't expect. We really needed to create an IT environment that worked regardless of what tools the user brought to the table. We certainly weren’t going to train them on new devices or systems.
• Cheap is good. In many cases, it must be very cheap, because we’re dealing with what is in many regards a disposable system. Most of what we built in my ‘space’, at least, will cease to exist in a few more days or weeks. We’ll be rebuilding an environment for Sochi late next year, but we’ll be using all new technologies—including SharePoint 2013.
• It must be easy to support on the IT side. We’re extremely busy and we can’t be spending time in the technologies. Once it’s up and running, it should be hands free, more or less.
• And flexible. It must be very flexible, so that it can adjust to new requirements and challenges in real time during the games.
• And scalable. It must scale quickly from the small handful of users who are working on the Games a few months ahead of time to the thousands who are working Games time.

What we are willing to sacrifice is “slickness.” Sexy, slick UIs, for example, aren't necessary. As you’ll see in upcoming columns, we invested very little in branding our intranet, because branding doesn’t directly support any of those requirements.

From those requirements, we built a SharePoint 2010-based intranet to support the collaboration, document-publication, and business process automation needs of the team. I’ll be sharing key points of that with you over coming weeks.

The reality is, nothing we do is particularly unique to the Olympics. Every business needs these things. The pressure cooker within which we build, deploy, support, and evolve the solutions is unique, and leads to a great learning experience that I will share with you.

I’d like to wrap up this week’s commentary with a well-deserved set of shout-outs to a couple of companies, products, and folks who were exceptionally helpful to us.
 

First, Microsoft itself. A few of the product team members helped us out at key moments and, of course, without them we wouldn’t have SharePoint in the first place! We were really hoping to leverage Office 365, but with the Preview beginning the week of Opening Ceremonies, it was just a bit too late for us. We’ll certainly be moving workloads into the cloud for Sochi.
 

Second, AvePoint. As many of you know, I had served as the Chief SharePoint Evangelist for AvePoint and the fantastic team there helped me get DocAve 6 in place at NBC Olympics to ensure our environment was managed, optimized, and backed up successfully. I had hoped to leverage the full capabilites of AvePoint’s new Governance Automation suite and its Office 365 integration, but due to my own time limitations I didn’t, this time around, but am already planning for it in Sochi. We’ve used DocAve for every Games since Beijing. In Beijing, we actually had a real data-loss problem, and DocAve’s granular restore saved the day. While those stories are fun, it’s much more comforting when the environment “just works,” as it did in London, and we can sleep confidently for three hours each night, knowing that the infrastructure is protected. I’m very proud to have worked for AvePoint and to have the opportunity to put it in place at NBC Olympics, where my own success depended on DocAve.
 

On the more “user-facing” side, you can imagine that there are a lot of great photos that we take at the games. There was a desire to share some of those photos with the team on the intranet, so we wanted an “image rotator” for several different scenarios. I put out a tweet and two teams responded in heroic fashion, providing me two image rotators that had slightly different features and capabilities, both of which we used.
 

The SharePoint911 team at Rackspace provided me the image rotator they use on their public website’s home page. They developed the rotator to demonstrate how an open-source jquery rotator could be used in SharePoint, refactored to pull data and photos from a SharePoint list. The demonstration became popular with their clients so they rolled it into reusable solution that was easy to deploy.
 

Kyle Davis also responded to my tweet-for-help with an image rotator that did some great things with photo resizing, which supported other scenarios where we weren’t able to manage the size, rotation, or aspect ratio of photos that were being uploaded into the rotation.
 

I have had the pleasure to work with Shane Young, Todd Klindt, Randy Drisgill and John Ross at SharePoint 911 in the past, and this was my first opportunity to work with Kyle. I was exceptionally grateful to all of them for helping me make a “win” with something that made a nice splash on our intranet! Numerous other folks came to my aid on specific parts of our effort, and I’ll shout out to them as I share those solutions with you. Thanks to everyone who helped, and who followed along on this journey. I look forward to diving into the technical details about what we did, starting next week! Stay tuned! (Read "SharePoint Intranet: How We Did It: Part 2.")
 

PS: for those of you junkies who didn’t get quite enough Olympics, NBC Sports is re-running highlights of the competitions for several weeks. Check out the schedule here--be sure to select your time zone to see the appropriate schedule.