In media technology in 2011, most of us are in the business of changing people’s behaviors on a grand scale—an objective rarely accomplished alone. Companies like Myspace, Facebook, and Twitter collectively created a paradigm shift in the way people consume and interact with media. Changing behavior so drastically was critical to making all these services more valuable than each would be by themselves. Myspace was one of the first to lay out the concept of the “Profile,” while Facebook was the first to really get photosharing right, while Twitter took the status update to the next level. Today, of course, Myspace has fallen by the wayside, but both Facebook and Twitter are benefiting tremendously from their collective efforts in getting people comfortable with profiles, photosharing, and status updates. And they are working together: “Post Tweets to Facebook,” “Share your Farmville score on Twitter,” etc. The same sort of “co-opetition” is true for many other emerging media industries.

Take music streaming companies (Spotify, Pandora, Grooveshark, Rdio, etc.), for instance. It is in their collective interest to get people out of the behavior of purchasing songs/albums, and comfortable with the concept of music streaming, the revenue/pricing models of these companies, and the different behavior it takes to use a music streaming service versus just purchasing music. Whether they like it or not, they are working together to create an entirely new way of consuming music.

In the business I work in (display advertising technology), consider the Demand Side Platform (DSP). This term was only born in 2009. And like a lot of media/ad companies, there were a lot of fast movers/competitors. But they worked together to create a market on a massive scale. When the DSP first came about, media buyers were accustomed to spending on less sophisticated and less technologically advanced ad networks. But after the term DSP was popping up just about everywhere due to the number of players in the market, media buyers changed their behavior and started planning their buys to include DSPs. And as these businesses and partnerships have matured, it appears that the DSP is here to stay because they provide media buyers a very valuable service.

In the above examples, it does seem that the whole is greater than the sum of its parts. In media technology, we are tasked with the tremendous challenge of scaling both the chicken and the egg. As daunting as that is, it is glorious to know that there are other people, on other payrolls, working just as hard as you are to create that market.

I’m sure there are countless examples of businesses where these ideas don’t apply (very unique and sensitive intellectual property, for instance—like that story about Thomas Edison patenting the light bulb an hour before another guy). But for media startups, my limited perspective tells me that accepting and even embracing your competitors will serve you well more often than not.

I ask the question in relation to this blog because we have seen a healthy market emerge trading digital ad inventory. I get why it works this way in practice, but I am more interested in the theory here. From an economic point of view, why do we have DSPs, SSPs, Exchanges, etc? Theoretically, why has the complex display ad ecosystem emerged? Is it here to stay like pork belly futures, or is it simply the product of a confused market, limited by technology, but unbounded by optimism in the short term?

To me, it seems too complex to be sustainable. But that’s why I want to understand the theory a little bit better. What do you think?

SAY Media’s Troy Young on the future of online advertising.  Brand APIs, Conversation Management Platforms, and Content Delivery Engines.

To me, it just seems like these guys are just delivering a new kind of coupon that wasn’t possible before crowd-sourcing was contemplated.  In other words, they are first to market.  I’m not saying this is an insignificant accomplishment by any means.  But I really do believe that it won’t be tough for new entrants (more local, more segmented networks) to compete in a few months.  These new entrants will quickly eat into Groupon’s margins and reach.

I think that all the hype for Groupon is based on revenue growth and margins.  I must admit that this is refreshing to see because so many of these internet startups are valued so expensively before they make any real money.  But I just don’t see their hockey stick growth curve continuing too much longer.  Unlike Facebook, for instance, they haven’t built anything that can’t easily be replicated.

And it’s my gut feeling, there’s nothing special about Groupon in the long term.

However, advertising has a lot of enemies.  People insult it all the time.  “I hate this commercial,” “This pop up is so annoying,” and “This site looks so trashy with all of these belly fat and teeth whitening banners.”  These are all valid concerns that advertising technology hopes to appease by delivering a higher degree of (1) ad relevancy and (2) ad/content integration.

Relevancy

Hulu.com has developed a great method of seeking greater relevancy:

By allowing the viewer to decide whether or not the ad is relevant is a surprisingly innovative way of accomplishing relevancy.  A lot of ad tech tries to target ads to consumers by algorithmically predicting their interests and behaviors.  But people are fickle and unpredictable, and maybe Hulu is onto something with this simple and elegant model.  And it politely requests the viewer to think twice about the ad and interact with it which seems like a much more valuable ad space than an ad on television.

Further, new types of media like Twitter and Facebook open a lot of possibilities for superior ad relevancy.  Never before have people been revealing so much information about themselves “publicly,” which–in theory at least–dramatically increases the odds of advertisers delivering consumers relevant advertising.

Integration

A pop up would be an example of an ad with terrible content integration.  A few emerging advertising technologies which I believe accomplish a greater degree of ad/content integration include Vibrant and Image Space Media.

Vibrant (and a few other companies) serves in text advertising:

When you scroll over a word of interest, an ad will pop up relevant to that word.  I agree that this can be a little annoying if you aren’t expecting it.  But now that I am used to seeing these ads around the web, I only scroll over terms that interest me.  Like Hulu’s model, I think this is a pretty elegant way of allowing the reader to choose the type of ads he or she would like to see.

Image Space Media has a similar technology that it applies to images:

When you scroll over an image of interest, you will see an ad relevant to the image.  I think this is awesome technology because it doesn’t cover up any text and because images seem like they would draw greater interest than text.  The only problem with this technology is that it seems to cause pages to load slower (most online display advertising does this to some degree).

In Conclusion

Much of ad spend on traditional media goes to waste.  Most commercials we see on TV are totally irrelevant to us.  And oftentimes we change the channel when commercials come on because they have terrible integration with content.  Greater degrees of relevancy and ad/content integration means a better experience for media consumers.  What it means for the advertising industry is more valuable advertising space.  With annual US ad spend of at least $100 Billion, we are talking about a huge opportunity for advertising technologists.

Big online publishers fill their ad inventory a couple different ways.  First and foremost, they sell it directly to advertisers.  Direct sales generally deliver the highest CPM rates, the highest margins, and the most relevant ad/content pairings.  A good example of a direct sale would be Nike buying 10 Million ad impressions on ESPN.com through an ESPN.com sales rep for a $10 CPM rate (a $100,000 buy).  Because ad inventory levels are unpredictable, however (because they fluctuate with traffic levels), most websites work with some type of ad yield optimizer to fill the rest of it not taken up by direct sales.

The job of a yield optimizer is to maximize the return from a website. And that means using sophisticated algorithms to retrieve and serve the highest paying, most relevant ads available each time a page loads.  It’s easy to do this in the Nike/ESPN example.  But when looking at the scale and value of remnant ad inventory on the internet, it isn’t economical to hand pick campaigns for specific sites.  Ad optimizers and their emerging technologies (such as RTB and behavioral targeting) are required to do this job on a massive scale pairing hundreds of different advertisers with hundreds of thousands of different websites.

The economic situation that I described in the first paragraph (few direct sales + a lot of ad inventory) means that ad optimizers are seeing a lot more business this month.  And in an industry with so many players, in a period of such hyper change and growth, this is the optimizers’ time to shine.

Who will step up to the plate and deliver the best performance?  I have a couple ideas, and I will be interested to see who is still around next January.  But in any case, in the online display advertising business, December is the month of the salespeople while January is the month of the technologists.

But what bet are they making?

The only way I see Facebook making money is on advertising.  To me, Facebook is just like a TV channel or any online publication.  I use Facebook to consume media.  What is different, obviously, is that the media is about my friends rather than famous people or broad culture.  But like most media, my life wouldn’t change significantly without Facebook–just like it wouldn’t change without ESPN, CNN, or the Discovery Channel.

Can Facebook deliver more targeted advertising than TV (or even Google) with all of the info it has about connections, preferences, and influence?  This is a pretty cool idea.  But more and more sophisticated ad targeting algorithms seems like a diminishing margin game.  People seem to be too irrational and unpredictable to ever completely conform to a mathematical equation.  Is this game diminishing before FB is worth several multiples of $50 billion though?  I don’t know.

Per Crunchbase, Facebook has 500 million users.  Let’s say each of these users views 5 pages per day on Facebook.  When I just loaded my homepage, I found 4 ads (screenshot at right), that don’t seem too relevant to my interests.  But let’s assume Facebook has great advertisers and their targeting abilities will improve dramatically in the near future which will include “Likes” by friends and anything else you can imagine.  And let’s say that Facebook gets the highest CPMs I can imagine for these ads: $2.00 each.  Thus, my very rough valuation of Facebook follows:

(500 million users x 5 pageviews/day x 365 days x 4 ads/page x $2.00 CPM)/1000=$7.3 billion annual revenues.

If Facebook has 1000 employees (per Crunchbase), and they each make a very generous $1 million/year ($1 billion total) and other operating expenses/overhead are another $1 billion, that leaves $5.3 billion left for “profit.”  Taking this number over 5 years and assuming Facebook’s growth will continue, the $50 billion valuation seems about right.  But this is obviously very hypothetical, and I haven’t done enough research to support the figures in the calculation.  The average of 5 pageviews per day may be high, and the CPMs are likely pretty high at this point too.  Those are the main two variables in this equation that have the potential to make a big difference.

The other variable that has an impact on this valuation is innovation.  What else can Facebook do?  Could they be a more closed off, social version of Google?  Will we consume all of our content including music, movies, and news through something like Facebook?

What do you think?  Is Facebook worth $50 billion?  What ways do you see us using Facebook in the future?  Thanks for reading and any insight you can provide!

The fill rate is the percentage of the time an ad is displayed. And fill rate is most often not 100%. A publisher might hear that an ad opportunity is worth $2 CPM. But if fill rate is left out of the conversation, this figure doesn’t mean anything. If the $2 ad only fills 10% of the time a page is viewed, the publisher is going to make only $0.20 per thousand page views (10% of $2.00).

Most often, a particular ad space is populated with multiple ads that rotate in and out all at different CPM rates. A lower CPM rate usually means a higher fill rate, and vice versa. Efficient online advertisers rotate the ads in and out to insure maximum possible effective (average) fill rates and CPMs per ad space on a website. Effective fill rate x effective CPM over a period of time equals revenue of a particular ad space on a website. Managing/balancing these rates can be very challenging and burdensome for an individual publisher lacking in time and resources. That’s why it’s a good idea for smaller publishers to work with ad optimizers or for larger publisher to have their own ad operations team.

If you are a small publisher without an ad operations team, choosing an appropriate ad network or optimizer is an important thing to consider. There are many variables that factor into ad performance, and some content simply monetizes better than other content. Usually, it is good idea to run a test with any new advertising provider before locking yourself into any contracts.

In conclusion, asking about CPM rates when evaluating an advertising opportunity on your site isn’t a very good question by itself. In order to actually project the performance of a particular advertisement, there are a lot of variables at play, most notably CPM and fill rate.

The above graphic from Terence Kawaja at Luma Partners represents the vast universe of players in the display advertising market. In one way or another, every company in this image is vying for a piece of the current $60 Billion annual ad spend, and trying to bring more online.

I’m not sure how many logos are on the slide, but I’m pretty sure that it’s too many.  Many of the smaller, less focused companies are being propped up by ambitious and idealistic venture money.  And many of the best small companies will likely soon be gobbled up by the bigger names on the chart.

It’s clear that the industry needs a number of players between the advertiser and the media.  These players most definitely include creatives and ad serving technology.  These players might also include media buying companies, aggregators, optimizers, and many of the other buzz words you see grouping these logos above.  These players probably don’t include companies without a technical advantage, advertiser relationship, or publisher relationship.  In other words, the players on the left are the creatives and agencies–they have the strongest advertiser relationships.  The players on the right do the ad serving themselves and they have the strongest publisher relationships.  Everyone else in the middle is fighting to produce superior technology to scale and get a piece of the ad spend flowing from left to right.  It’s a really interesting graphic and it will be even more interesting to see it as the months go by.

Thanks for reading.

Media6Degrees CEO Tom Phillips explains the difference between behavioral targeting and social targeting in online advertising.  Really interesting perspective from one of the industry leaders.

The video loads very slowly for some reason, so it may be easier to pause it for a while before viewing in full.