media landscaping

Internet Math Applied to Video and Social Networks

One of the things I enjoy doing is reverse engineering bogus numbers thrown around by supposed 'experts' or start ups.  Online advertising is so accountable and data is so prevalent that these wild claims are not to hard to call BS on.  It seems that others enjoy this as much as me...here are 2 great examples:

VideoNuze published a very insightful breakdown of instream video advertising revenues on the web and points out some major holes in analyst forecasts.  According to Mugs's math, YouTube (34% of total web video views) can only deliver about $22.6M worth of instream ads this year and the big four broadcast networks won't break $120M combined.  Analysts have predicted up to $1.3B in online video revenue...so where is the rest of the $1.16B coming from?  Conclusion: analysts full of crap.

On another note, Mashable invents the Screw You Coefficient used by social networks to determine if a new feature should be launched or not.

It totally works.  Check out the post for more details.

Pubmatic CPM report very misleading

This report by Pubmatic was passed around the office today.  It says that, according to data gathered from the 3,000 sites that Pubmatic manages ad inventory for, small sites earn a higher eCPM's than the larger ones.  Many people, including commenters on techcrunch, seem to think that this is because smaller sites have more targeted demographics and therefore get higher CPM's.  This is only partially true.  The main reason is depth and frequency of use.

The information Pubmatic is offering is only the ad network 'unsold' inventory and not ads the sites sold themselves or through other channels.  One thing about ad networks that you may not know is that they highly value the 'early' part of a site experience and pay far less for the 'later' impressions.

Translated: Consider 2 sites that are exactly equal except one has an average pv per unique of 5 pv's and the other 10 pv's.  Site one with only 5 pv avg will earn a much higher eCPM than the site with 10.  Ad networks earn (and therefore pay) less on the later impressions. 

I am making an assumption here, but I think that the larger sites Pubmatic manages have a deeper and more frequent readership than the smaller ones do...especially if any of the larger sites are social networks.  That is why these numbers are so off.

Are the "Influentials" Dead?

Duncan Watts, a network-theory scientist who now works at Yahoo, has performed many experiments that disprove the true influence of 'The Influentials'.  An article in Fast Company, Is the Tipping Point Toast, lays out Duncan's beliefs that all the money spent by marketers to target Influentials has been wasted.

Watts believes [...] a trend's success depends not on the person who starts it, but on how susceptible the society is overall to the trend--not how persuasive the early adopter is, but whether everyone else is easily persuaded. And in fact, when Watts tweaked his model to increase everyone's odds of being infected, the number of trends skyrocketed.

So rather than target influentials, marketers should be looking for the easily influenced?   That sounds like fun.

(via Ad Lab)

The Future of Television Advertising

A panel at the Future TV conference in New York discussed what TV advertising will look like in the future and it sounds strangely familiar to the world of internet display advertising today.  As technology evolves to allow targeting down to the household level.

By aggregating the disparate local cable advertising in ways that deliver meaningful sub-segments of viewers, Steib [Mike Steib, director of Google TV Ads] said Google TV ads program is able to create audience mixes that likely would have higher advertising value than their remnant avails currently have on their own, and which theoretically could compete in value with some of the TV industry's most premium network TV inventory.

"There's all these opportunities to drive sell out," he said, "much, much closer to 100% and to take the CPMs up significantly when you start matching the right advertising with the right audiences."

Joan Gillman, president-media sales at Time Warner Cable, and Barry Frey, senior vice president-advanced platform sales at Cablevision, concurred with Steib's vision, though they didn't necessarily think it would be the exclusive province of Google.

"We can give the data, that measurement is there, but the actual customer data is at the operator level," Gillman said, portraying a vision for aggregated TV advertising networks that are not unlike the kind that are becoming increasingly popular in the online world, which are based more on the behavioral targeting of consumers than traditional demographic targets.

The initial vision for this is aggregation of "long tail" content across multiple channels and selling them static as one package, which is nothing like dynamic behavioral targeting at all.   Its a start but is still very far away.

Real TV behavioral targeting would be serving ads to viewers based on what they have watched in the past...not the content they are watching now.  The cable-box would need to have a memory and be able to transmit viewing data back to the ad providers, who then send an ad customized to the viewership of that TV.

The MediaPost article about this panel, Future Of TV Ad Market: A Lot Like Online's -- Ad Networks, Behavioral Targeting, Etc. is interesting but left me wanting more.  TV still has a long way to go to catch up to the efficiency of online advertising.

Movies on a cell phone

Mobile evangelists love talking about video on mobile devices and how great it is.  I get that pulling up your favorite youtube video at a bar to show your friends is a great use of this, but David Lynch obviously doesn't think this will be so great for watching a feature film.  Enjoy.

Thanks to Ken 'The Samurai' B for forwarding this to me.

Network Journalism Summit Morning Update

I took the 6am flight to NYC and made it just in time for the start of the Network Journalism Summit.  The first panel was about how traditional newspapers are interacting with citizen journalists by creating sites that give them a voice.  The most interesting point to me is that most of these organizations reverse-publish content from the community participants in the print version of the paper.  They all say that this strategy is the main carrot that drives participation in the communities.

John Wilpers of BostonNow.com said that of the 500 community bloggers they have, at least half of them have been published in the paper itself.  He even said that they are looking into ways to compensate these bloggers monetarily, based on traffic, clicks or some other measure of their impact.

My panel went well...I think...but we will have to see what the people say.  You can follow the conference through the tag Netj.  Here is the technorati link that I am tracking the conversation with.

More to come later...

See you at the Networked Journalism Summit!

I am really psyched to be a part of tomorrow's Networked Journalism Summit at CUNY Graduate School of Journalism.  It is organized by Jeff Jarvis and made possible by a grant from the MacArthur Foundation.  The vision for the event is:

bringing together the best practices and practitioners in collaborative, pro-am journalism

and I was invited to talk about our learnings from the WPNI Blogroll Program.  You can read a summary of my pre-interview with David Cohn here.

I'll post some afterthoughts and maybe even live-blog if possible.  We'll see what happens.

Amazing Image Technology and the Newspaper

I just came across this video from a recent TED Conference.  The technology is completely mind-blowing, but I was particularly fascinated with how this could impact the future of 'electronic editions' of newspapers and magazines. 

I still don't think that people want to read their paper this way, but with infinite resolution like this I could end up changing my mind.

Wanted: Widget Advertising

Brands that are truly embracing todays world of social media are experimenting with widgets because they know that people will evangelize and spread their love of a brand if given the tools to do so.  One good example is the Purina Pet Weather Widget.

The one thing that I am seeing very little of is actual advertising in widgets.  With the popularity of widgets growing like wildfire, I suspect that media companies who are creating these widgets will start selling a portion of the space to advertisers.   Here is the first example that I have found:

Notice the Lenovo Sponsorship at the top.

They also produce some video highlight widgets with ads from Sprite in their Video Blog, but the code didn't work right.  Here is an image of it on the left.

I think that there will be a few challenges for this type of advertising.  First of all is size.  Widgets are small to begin with, so adding advertisements runs the risk of seriously hurting the experience.

The second challenge will be my Theory of Widget Advertising Elasticity and its affect on widget adoption.  As the % of space allocated to advertising  increases, it will decrease the willingness of publishers to place that widget on their site.  This is a very important thing for media properties to consider and I believe it is the main reason we aren't seeing much of it yet.

Due to the fact that the NBA is a very premium content owner and you cannot get this content any other way, I think they have a very high widget ad elasticity.  Therefore, these ads are probably not impacting their adoption by bloggers and other publishers.

If you notice any other widgets with ads like this, let me know.  I want to keep a close eye on this topic. 

More to come...

Testing out the new 'share' video syndication tool on washingtonpost.com