Tag Archives: user-generated-content

Analyzing the CBS YouTube Stats

Today, CBS issued a press release celebrating their first month on YouTube. I first heard of the press release on the CinemaTech blog.

Coincidentally, I was just writing a Ruby program to access YouTube data for a project I’m working on with some friends at UC Berkeley… so I decided to put it to the test against CBS’s press release data.

According to the CBS press release, the Top 15 CBS videos watched this month (as of November 17) on YouTube were (Total Views in parentheses):

1. NCIS/Cat Fight (1,603,364)
2. Letterman/ Borat Meets David Letterman (1,057,180)
3. Early Show/ Borat Vs Harry Smith (969,391)
4. Letterman/ Bush is drinking again (698,806)
5. Letterman/ Message About February for Bush (524,697)
6. CBS Evening News/ Michael J Fox Talks to Katie Couric (465,563)
7. CSTV/USC Cheerleaders: The Song Girls (374,623)
8. CSTV/ A Field of Dreams for Judy Coffman (358,572)
9. Letterman/George W. Bush Fakin’ It (357,213)
10. Letterman/ Dave and Bill O’Reilly (352,747)
11. Letterman/The Guy Who Swears At Dave (316,258)
12. Ferguson/ Fun is Dangerous (314,093)
13. Letterman/Do Maggots Go With Scorpion (278,283)
14. Ferguson/ Bush visits the Un-Late Late Show (221,462)
15. Ferguson/Bad Kerry (219,556)

According to my stats, the Top 15 CBS videos as of today (November 21) are (I’ve added the Upload Date at the end of each video):

1. NCIS Cat Fight (1,626,323) [10/17/2006]
2. Borat Meets David Letterman (1,420,773) [10/30/2006]
3. The K-Fed/Britney Sex Tape (1,152,283) [11/9/2006]
4. Borat wrestles Harry Smith (1,038,996) [11/1/2006]
5. Rumsfeld Gets Cute At The Podium (826,849) [11/9/2006]
6. Britney Spears Surprises Dave (815,027) [11/7/2006]
7. Bush is drinking again! (Late Late Show) (728,940) [10/16/2006]
8. Rumsfeld Gets Cute At The Podium (extended version) (668,323) [11/13/2006]
9. a kiss is not a kiss… (610,691) [11/3/2006]
10. Letterman’s Important Message About February for George Bush (550,967) [10/26/2006]
11. Cheerleader Whacked by Leprechaun (536,502) [11/13/2006]
12. EXCLUSIVE: Michael J. Fox Talks To Katie Couric re Rush L (473,574) [10/27/2006]
13. Playstation 3: It’s Just so Old and Outdated (414,019) [11/17/2006]
14. Kirstie Alley’s Bikini Shock (388,452) [11/20/2006]
15. USC Cheerleaders (The Song Girls): CSTV (384,265) [9/29/2006]

It’s a very different list, with eight newcomers (shown in red). Now, I have no practical way of obtaining the data for Nov. 17, so I have to work with what I have. Either CBS doesn’t want to advertise certain shows or there’s some very interesting viral growth going on here.

The newcomers where all uploaded during November, so they got quite popular very fast. Now, I don’t think CBS has some hidden agenda to hide K-Fed/Britney shows and Rumsfeld sketches (they seem quite happy to show three Bush clips on their Top 15). On the contrary, this gives me an opportunity to do some quite interesting analysis (at least until I gather more complete data on my own).

On November 17, according to CBS, a show needed about 219,000 total views to make the Top 15. That means that on 4 days, eight CBS shows gathered enough views to break into the Top 15:

At #3, the K-Fed/Britney Sex Tape (on Craig Ferguson), must have garnered almost 1,000,000 views on those four days. At #5 and #6, Rumsfeld and Britney needed about 600,000 views on those four days. At numbers 8 and 9, Rumsfeld again, and The Adventures of Old Christine, entertained close to half-a-million viewers.

Let’s list the newcomers, along with the number of viewers that watched them between November 18 (after CBS’s Top 15) and November 21 (my Top 15):

3. The K-Fed/Britney Sex Tape 932,727 Viewers [11/9/2006]
5. Rumsfeld Gets Cute At The Podium 607,293 Viewers [11/9/2006]
6. Britney Spears Surprises Dave 595,471 Viewers [11/7/2006]
8. Rumsfeld Gets Cute At The Podium (extended version) 448,767 Viewers [11/13/2006]
9. a kiss is not a kiss… 391,135 Viewers [11/3/2006]
11. Cheerleader Whacked by Leprechaun 316,946 Viewers [11/13/2006]
13. Playstation 3: It’s Just so Old and Outdated 194,463 Viewers [11/17/2006]
14. Kirstie Alley’s Bikini Shock 168,896 Viewers [11/20/2006]

Those are very interesting numbers for a four day period.

I also drew a couple of graphs to try and understand my CBS data better.

The first one plots CBS’s 328 YouTube videos, charting Total Views by Upload Date. Though a whole bunch of videos are down in the under 200K view range, there might just be a pattern forming where newer videos are gathering more views. Eventually I’ll be able to plot views over time.

CBS: Total_Viewers vs. Upload_Date

The next graph, plots Total Views for each video against the Average User Rating they received. On YouTube users may rate a video on a 1-5 scale. A “0” means the video has not been rated by any viewers. These graph shows that most CBS videos receive a rating of 2.5 or more, and that the more Views a video gets, the higher its average rating. Nothing extraordinary here.

CBS: Total_Viewers vs. Average_User_Rating

The final graph for today plots Average User Rating vs. Upload Date. Again, nothing clear but a very slight pattern might be forming where newer videos are getting higher ratings. The other funny pattern obvious from these Upload Date graphs is that CBS does not upload any videos on weekends – as evidenced by the small vertical gaps every five days.

CBS: Average_User_Rating vs. Upload_Date

So there you have it. A first attempt at analyzing CBS on YouTube. Expect more as time goes by and more data comes in.

Technorati Tags: , , , , , , , ,

Google Does Video Graphs

Well, it seems someone at Google is listening. Things can only get better.

I was checking out the videos at The WorldTV Internet Charts and clicked on one of the Google Videos. After watching the video I noticed that, just below the star ratings, it said “All time views: 712,721 »“. So you know I had to click on the little chevron symbol to see what lay beneath these numbers.

And Voilà!

Google Video Graphs

There they were… historic audience graphs for the last week, and a table showing the All time Views & Rank, as well as Yesterday’s Views & Rank, including how many were via email or embedded in blogs.

Hey, it’s not the Google TV Dashboard I wrote about, but it’s a start.

As a video producer or advertiser, I’d also like to know how many people watched the entire video and how long did the rest of the viewers watch. These graphs would look something like this (we could call it the “saddle-graph“):

Saddle Chart of View Completion

The name, of course, comes from the saddle shape of the graph. I haven’t seen one of these but I think it’s safe to say that while many viewers would stop watching a video at the beginning, those that kept watching would tend to stay until the end, given the time they had already invested in the video. That’s why you get the peak at 100%. In television, you tend to see graphs like these for shows that tell a story (and thus capture the viewer), while for other types of content you see more irregular graphs.

The Reports page for your uploaded videos still only shows a simple table with Page Views & Downloads. You can choose the time frame for the reports but that’s about all you get.

Page Views & Downloads at Google Video Advanced Reports

And the numbers you do get, don’t necessarily match the ones on the Video page. For example, my two videos of the Atlanta Aquarium show 1315 & 2873 Page Views on the Advanced Report page, but if I go to the actual link and watch the video, I see 1330 & 2957 respective All Time Views. Not exactly dependable, but I’m sure they will eventually get it right.

There’s more information at the Google Video Blog, including a great new way to add comments that link to exact points on a video (so if you think the funny bits begin after the first three-and-a-half boring minutes, you can link straight to them).

Want more articles related to this post? Check these out:

The Advertiser’s Dilemma

Rethinking Ratings

Why Google Should Buy YouTube

Google Media

Technorati Tags: , , , , , , ,

Two Interesting Articles from Advertising Age

Advertising Age has a couple of interesting articles today:

Ad Age Digital

User-Generated Video Gets Its Own TV Channel

According to this article, Fame TV is a new channel dedicated entirely to consumer-generated videos, airing on the BskyB satellite network in the UK and Ireland. Users can upload their videos for a $3 fee -there’s also a deal with Revver for content- and these will be aired using an innovative format: nine boxes will be on the screen at all times, each one showing a different video. Each box has a code so that viewers can vote via SMS for their favorite show. In the case of Revver videos, Revver will pay a percentage of the SMS revenues to each video creator.

I’m assuming that videos getting lots of SMS votes, will be aired more often to create even more revenue. Judging from the success of shows like “America’s Funniest Videos,” the success of the idea seems guaranteed. It’ll make for a great time-waster and sure beats channel surfing.

It will initially lack advertising, though a good idea would be to run advertising in the center box (an even better idea would be to user-generated advertising, a la CurrentTV’s V-CAM (Viewer Created Ad Messages)).

Locally, this would be a great idea for networks to fill an hour or two of late night infomercial infested air time.

and the other article is:

Ad Age Media Works
Why TV Needs Commercial Ratings — Now

CBS‘ Chief Research Officer, David F. Poltrack, explains in great detail why it’d be smart for the television networks to measure ratings by the number of people watching commercials instead of the number of people watching the shows.

While this may sound counterintuitive at first, he gives a detailed explanation oh what’s happening with DVRs. Viewers are increasingly using DVRs to watch TV shows at their convenience. As I’ve previously mentioned here, network competition is moot when you can record both competing shows and watch them both at a later time.

When viewers have a DVR, more of them watch the shows after the fact than live. And about 40% of these viewers actually watch the commercials. Well, according to Poltrack, these eyeballs are not being counted, since the current system relies on live viewings (because advertisers automatically assume that DVR owners will fast-forward through their expensive commercials).

With the proper measuring system in place, viewers who watch the ads will be accounted for, while those who skip them won’t. Sounds fair to me, but go read the whole article… recommended.

Technorati Tags: , , , ,

The Advertiser’s Dilemma

The typical life of an internet video goes somewhat like this: someone uploads the video to a video sharing website (YouTube, Google Video, Metacafe, Dailymotion, your pick) and sends links to his friends telling them to watch the video. One of them thinks it’s funny and passes it around. Some other guy finds it on the website, writes a comment and shares it with his contact list. If the video is really funny, sometimes (not always) it will explode and become an internet hit (like the Star Wars kid or Pinky the Cat). Millions will watch it and pass it around. And eventually it’ll become old and die a natural death (sometimes to be resurrected further down the line).

As an advertiser, you’d want to identify these runaway hits before they become a success (thus minimizing the number of eyeballs lost to your message). As a content producer, you want to convince advertisers to buy space on your video, before you lose the advertising value of all those eyeballs.

So, how do you maximize your return on advertising on web videos?

Traditionally, we could say advertisers have it easy (though I’m aware how hard ad buying really is). Television networks have been around for a long time, have time tested products, experienced programmers deciding what gets on the air and when, and a captive audience. They also have a company (AGB Nielsen) that measures all these shows down to the minute, reporting on the age, sex, location and income level of the viewers.

The internet, however, presents a whole new set of unknowns. Most content delivery websites have no control over their content producers nor do they know who these producers are. There’s no experienced programmer deciding what gets showcased (CurrentTV does, but they have a different business model and approach to web video); instead, other users rate the videos according to their personal tastes (and with a little work this system can be gamed very easily). Finally, there’s no real measuring going on (I’ve written about this particular issue here, here and here). Most websites simply tell you which video has been viewed the most, or ranked the highest (again, with highly suspect numbers). And what they know about their users is usually limited to their email address, what they’ve published, viewed or ranked and maybe an IP address that can suggest where they connect from (which used properly can be a very useful variables).

As an advertiser, you’d want to optimize your purchases (as opposed to buying ads on every conceivable video and hope one of them becomes an internet sensation). But by the time you can tell a video is a runaway hit, you’ve not only probably lost the majority of your potential audience (sort of like entering a pyramid scheme at the bottom), but you’ll also have to pay a premium to advertise on that now world famous video.

We need tools that can track the spread velocity of a video, their viralness, so to speak. We also need to define new demographic variables, based not only on age / sex / location / income but more importantly on interests and social connectivity. When you have 14-yr olds playing online games against 30-yr olds, age and sex are no longer as relevant as what interests these people share.

Of course you can simply blanket every uploaded video with your advertising, but would you rather be that ad on every lousy home video, or that cool ad on the hilariously popular video-du-jour?

Technorati Tags: , , , , ,

An Informal Chat with Second Life Creator Philip Rosedale

The first Forbes MEET Conference was a blast. The panels were great and the people attending were even better. The conference is more than worth the price of admission for the networking opportunities alone.

On day one I had the good fortune of sharing a lunch table with Chad Hurley and Chris Maxcy of YouTube, Neil Kleinman, Dean of the College of Media and Communication at UArts, Janet De Vries (Director of the President’s Office at UArts) and Philip Rosedale, the creator of Second Life.

Neil is doing some wonderful stuff over at UArts, including getting the College of Media Communications online and the students interested in all that the new media technologies have to offer. It would be great to see art students presenting their creations in a Second Life exhibit or theatre majors opening their shows on Second Life‘s version of Broadway.

Second Life, for those unfamiliar with it, is -to quote their website- “a 3-D virtual world entirely built and owned by its residents.” You travel to Second Life’s world by downloading a small program from their website and registering your “avatar” or virtual representation. Once inside the game, you decide whether you want to be male, female or something entirely different, like a bug. Choose your clothing, change your looks… anything you want. You can also buy land to build a house on (or a cabin, high-rise or space station). Set-up a clothing store and sell your shirts to other Second Life residents. Anything is possible here…

Many companies (Sun, Pontiac) have set-up shop inside Second Life, several residents are making good money buying and selling stuff (houses, cars, condos), and even Reuters has sent one of its reporters full time into Second Life – which I find mind-boglingly wonderful.

I don’t really have all that free time at the moment, even if I could build a lucrative business inside Second Life. So I asked Phil when will I be able to send my Second Life avatar to do work for me and report back at the end of the day. Phil laughed… this simply isn’t possible, yet. But, wouldn’t that be great? I guess that, eventually, avatars will be programmable to follow certain scripts and maybe even interact on their own with other avatars. Maybe then I could create for-hire armies (ok, or gophers) inside Second Life and rent their services out for a nice sum.

On a more serious note, Phil did mention that great things are coming to Second Life’s audio/video functions, including 3-D audio, which would allow you to determine where a sound is coming from (much like you do in games like Counter-Strike).

Technorati Tags: ,

Conference Over

The MEET 2006 Conference is now over. Live blogging was neither practical nor possible, but I will be posting my comments over the next few days. My laptop is acting up, so bear with me while I fix it. I had the chance to have lunch with Philip Rosedale, the creator of Second Life, as well as the YouTube gang (Chad Hurley and Chris Maxcy).

Michael Eisner interviewing Barry Diller was both hilarious and insightful. So was Brad Grey’s interview. The conference’s panel format was great, as it lends itself for great conversation, insights and personal dynamics. It’s refreshing to come out of a business conference without having seen a single powerpoint presentation.

Rethinking Ratings

Summary: An analysis of current television ratings methods, why they’re inappropriate for the timeless internet and digital video recorder era, and suggestions for improving them.

Traditional television ratings reports let TV executives and analysts study the behavior of a particular show or series, displaying the number of viewers each show had, broken down by demographic targets. This allows the television industry to determine which show won a particular time slot (e.g., Friday 9pm to 10pm), how it performed among a particular demographic (e.g., Males 18-34yrs) and how it has evolved (in the case of serials) over time (e.g., Are there more or less people watching it).

But what happens when viewers can watch any show at any time? When viewers don’t have to choose one show over another on a rival network? What happens when you can’t tell for sure who your viewers are?

The internet and TiVos give the viewer unprecedented freedom over when, where and what to watch. Soon it won’t be possible to tell for sure how many people are watching any given show, using traditional ratings tools such as AGB/Nielsen‘s Peoplemeters. Programming executives won’t have to worry about what the rival networks are showing at the same time as their new hit show. And ratings analysts won’t be able to track a new series’ behavior by simply looking at how each episode did on its air date.

Two hit shows going head-to-head on rival networks? Not a problem: watch one and record the other for later viewing (or get it from the Net). Missed last week’s premiere episode? No problem there either: watch it online, download it off bit torrent or pay for it on iTunes. Some of this you can easily track, but some you can’t.

Analysts will need to track each episode over time and then track the series as a whole. A VERY SIMPLIFIED graphic might look something like this, with a running total for each episode over the time of the series:

VERY simple ratings graphic

Any viewer can watch any episode from its air date to the end of the series (and beyond). This allows viewers to catch-up after the series has started or to catch any episode they may have missed. Of course, the whole concept of missing an episode disappears in the TiVo/Internet model. But in addition to tracking how many times a particular episode was watched or downloaded, you should also be tracking what’s happening with the rest of the show’s internet presence. Are viewers reading the characters’ blogs? Are they discussing the show in the forums? Are they setting up fan websites? Linking to the Myspace profiles? Uploading mashups of show clips? Not only must you track the show’s behavior over time and over several distribution methods, but you must also track and measure the user experience surrounding the show.

And finally, how do you solve the demographic problem: if you don’t know who your viewers are, how do you target them? The answer is both simple and complex. I believe that traditional demographic targets are on the way out. Social networks and special interest groups are the new targets… and these are much easier to track via the Internet than the old ones. You may not be able to tell whether a particular viewer is male or female, young or old, wealthy or not, but you can tell what news s/he reads, what games s/he plays and which people s/he hangs out with (to a certain degree, of course). One minor detail… you can’t (or shouldn’t) add apples and oranges. Traditional television ratings data categorizes viewers by demographic targets such as age, sex, location and income (because someone takes the time to visit each household in the sample and verify this information). And whereas traditional ratings analysis has always relied on a sample set of data subjects, internet traffic and behavior analysis has always examined the whole dataset. Eventually it shouldn’t be too hard to homogenize both sets of data, either by linking traditional television viewers to their online behaviors, or simply by expanding their interviews to include enough data to categorize them.

Currently, Google and YouTube limit their video data to a traditional web-traffic analysis mindset: most viewed, most recent, most subscribed. Coming from an Internet world, they fail to see the need (or maybe even the possibility) of better, more detailed reports (Yes, it could also be that they keep these reports hidden from the outside world).

As for me, I’d love to know how the most watched videos on YouTube evolved over time. Have they peaked? Are they growing? Who watches them? How about a Google Finance like chart, linking views to blog/news mentions? Which video has been linked-to the most (this one is actually on YouTube)? Which videos have been dugg and how many diggs did they get? Actually… I’d just love to work there and get it done myself!

Off to the MEET 2006 Forbes Conference

I’m off to Forbes’ MEET 2006 Conference (tomorrow and Wednesday at the Beverly Hills Hotel). The conference theme is:

Reaping Riches in the Media and Entertainment Revolution.

Check out the conference website for the agenda and list of speakers. I’m not sure what the blogging policy will be, but I’ll certainly try to post live if allowed.

After the conference I’ll be visiting San Francisco for a couple of days.

Leave a comment if you’re going…

CBS Launches YouTube Channel

CBS, one of the US leading television networks, has launched a YouTube channel. So far the content is limited to short clips from late night television, sports highlights, program promos and news items.

CBS Launches YouTube ChannelI’m not particularly impressed with the available content (no full length shows yet) but I really like the fact that CBS has taken this initiative. Late-night clips were already showing up on YouTube, so why not offer them straight from the source?

As of this writing, CBS’s YouTube channel has about one thousand subscribers and 33,000 views. As more -and better- content from traditional networks goes online, it’ll be interesting to see how they compare with user generated content. Will they reach the 37 million views attributed to smosh or surpass lonelygirl15‘s fifty thousand subscribers?

It’s great to see the traditional networks embrace their fears and venture online. CBS already has content distribution deals with iTunes (Lost is sold on iTunes for US$1.99 per episode), but I’m guessing they’ll release their shows for free on YouTube under a Google advertising supported model.

Interesting times, indeed.

Where are the Editors?

Fellow blogger John Allsopp (dog or higher) writes about the future of blog reading, and how we’ll go from reading blogs to reading single posts.

In my personal experience, that’s exactly how it goes. Although I subscribe to several blogs, I find myself not checking them very often, instead relying on Google, Digg or Del.icio.us to point me in the right direction. Of course there are still some industry blogs I check on a daily basis (GigaOM, TechCrunch, Mashable), but the rest of my reading (such as Allsopp’s article) is simply per-post.

Bloggers have many and varied interests and more than likely, not all of their posts will be of interest to me. Though you may read my blog because you’re interested in Digital Media strategies, that is no guarantee that you share my interest in Privacy, Security and Usability.

Which brings us to a new problem (or more accurately, a new version of an old problem): in a world of endless content, how do we quickly find the good bits?

Finding Content

While television has had programming experts choosing what and when to show (in addition to hundreds of specialty channels with even more specialized programmers) and newspapers have editors, in the online world we’ve had to rely on automatic digital aggregators (usually based on tags or keywords) or other users (most of whom we know nothing about) to choose the most relevant content.

Other services, such as Findory, look at your reading patterns in order to show you relevant information (as long as you read it through their interface). And though I’ve used Findory before, I haven’t yet been able to integrate it into my daily workflow (and I always get the feeling I’m missing out on some relevant item – I’m not quite sure why that is).

The problem with digital aggregators is that not everyone tags their content, there’s no tagging standard, and not all tagged content is good or even relevant. I’ve subscribed to Google Alerts and Technorati tags, but must compromise between general tags -and lots of false positives or irrelevant content- or very specific search terms -and thus missing out on some possibly relevant articles. 

On services like Digg it’s very easy for a group of users to control the system and get their content on the front page. Get a bunch of your friends to digg each other’s articles and you’ve instantly got a leg up on everyone else.

And though I mostly use Del.icio.us to search for my own bookmarked information, I’ve noticed its search results are usually quite relevant. I believe this has to do with users tagging content for their own future use -as opposed to tagging for the community- and do a better job with it.

Choosing Content

What we need is the online equivalent of editors. A trusted and accountable system to separate the good from the ugly. But, do we want them? Have we moved away from traditional media (from newspapers towards blogs) only to come back to a traditional model? Or is this a new, evolved model, where power remains in the reader’s hands?

The answer probably lies in a mixed system, borrowing the best of both worlds, much like TiVo has done. TiVos (or Digital Video Recorders) allow you to record your favorite shows and watch them at a later time. You’re no longer tied to a particular station’s offerings or timeslots. In a sense, you’re a programmer: you decide what is on and at what time. But, and this is important, you only get to choose from a pre-established pool of content. Yes, it may be great to watch Lost, Heroes and 24 back to back, even though they may be on competing timeslots or different days on broadcast TV, but you’re still picking your shows from what the network programmers think are the best of the best.

It will be interesting to see what happens when Apple‘s iTV comes out, or when Google finally decides to offer a Universal Video Recorder and you can choose your content from broadcast (chosen by programmers) and the Internet (chosen by you or some search / tagging / voting / aggregator service).

What do you think? Will we be letting editors choose our content? Or will we keep searching on our own for the best content? Where’s the middle ground? Leave a comment and let the world know what you think.