Adonomics™ Facebook Analytics and Developer Services

Dan Farber, a Google fanboy at the recent SNAP Summit, has clearly bought into Google’s “Do No Evil” line.

Dan shared the story of how at a dinner with a top technologist from Google on the night before the Summit, he was told that Google was about ready to be VERY OPEN with key parts of their API’s, data and platform including access to whatever subset of the World’s Social Graph that Orkut might represent. The contrast Dan Farber was trying to draw here was that Facebook was being VERY CLOSED about their much more significant subset of the World’’s Social Graph and that it was clear that Google understood and supported open web standards better than Facebook, and well, “Go Google, Boo Facebook!”

I considered asking Dan’s panel the question about how likely it would be for Google to open up something that actually made them money (e.g., their AdSense inventory that exists out on millions of individual web sites or the right hand Sponsored Link area of Google.com for which Microsoft and/or Yahoo would love to compete to serve the most lucrative ads for the web site owner’s benefit).  I knew the answer to this was, ”About as likely as iTunes opening up and supporting other music players besides the iPod — that is NOT AT ALL LIKELY.”

Software Capitalism 101

My actual question to Dan was about: “What dominant business had ever opened up to its competitors the core of their monetization engine?” was met with stumped silence from the panel.   One comment came about how Firefox (the ghost of Netscape’s proprietary browser) was openned up by AOL and it now has 200+ million users but this was a case of the losing browser opening up as a final act of desperation to somehow stay relevent.    So, it is the weak competitors who want an open standard not the strong competitors whose definition of OPEN is you are free to write to our OPEN platform.    In a related comment, Robert Scoble finally said that it was really just bloggers in the tech world (and competitors of facebook) that really want to have a WEB-WIDE OPEN ID STANDARD for accessing the user profile data inside of Facebook (and not end users who really don’t care that other companies who want to access Facebook’s data will have to do so by writing facebook apps).

Lookery’s  Big Idea and Facebook’s Nov. 6 Announcement

This got me thinking about what Facebook and/or Microsoft might be announcing on Nov. 6, 2007 in terms of their “Advertising Play.”   I had heard Scott Rafer at VideoEgg’s App-Camp and Todd Sawicki at the SNAP Summitt make the following points:

1. Facebook has great demographic data and horrible click-thru on their ads — because their users like to stay on Facebook, no matter how well targeted the ad is

2. Most web sites have no demographic data and nice click-thru on their ads — because their anonymous users are just passing through as they browse for information

3. Nirvana would be facebook’s demographic knowledge with a web publisher’s low site stickiness.

Lookery plans to help address this by taking facebook collected data and then enhancing the value of the ad inventory that is available out on the rest of the web. This is a GENIUS IDEA and one that I hope Lookery has found a way to patent because I believe that Facebook and Microsoft are likely to co-opt this idea on Nov. 6.

So, here is my prediction, based on a variety of comments from folks at facebook and from other ad network providers, about what the headline and big story will be on Nov. 6:

The Second Shoe Drops — Google Threatened as Facebook Announces Web-Wide Ad System

(AP — Palo Alto, CA)  After losing out to Microsoft in their recent attempt to prevent Facebook from staying a “Google-Free Zone,” Google’s own AdSense business is now under direct threat from Facebook and Microsoft.  The threat from Facebook, a technology company focused on helping its users communicate with each other, access information about their friends and share the most relevant information with others, has today announced “fbCash” which you can think of as a “Facebook-Enhanded, Open-Source, AdSense” software solution and service for web publishers that are currently locked into a Google-only AdSense solution. This technology works by using non-personally identifiable data that Facebook has on 50 million users (growing to 200 million by Dec. 2008) to selectively target the ads on a web publisher’s site in a manner that generates a higher click-thru and higher CPC-rates and higher revenue share for web publishers than Google’s proprietary, closed AdSense solution.

This is an out-growth of the Facebook Ad Network that was originally being designed to operate only within facebook apps. In this new system, facebook will turn over to an Open-Source group the ad monetization algorithms that will dynamically test Google AdSense ads against Facebook-enhanced Microsoft ads against Yahoo Ads, etc. to determine the particular ads that are best suited for a particular user of any given web site.  These monetization algorithms will take ad feeds from any dominant ad providers that are willing to provide a description of how much of the advertisers CPC ad payment is shared with the web publishers.  Individual web sites can also access this Facebook-enhanced user data to determine if a given user to their site should be offered a special coupon or shown a particular landing page (e.g., one that is best for a 25 year-old, woman from New York City vs. one that is best for a 13 year-old, boy from Cairo, Egypt).

Background on Facebook Ad Networks

Most developers that have tested advertising on their facebook apps have discovered that it is hard to make ends meet with a single ad provider. The typical experience is to start with Google AdSense. In a very short period of time, the developer figures out that Google AdSense (which is based on reading the text of the facebook canvas page and trying to put up a relevant ad) is ill-suited for Facebook for three reasons:

1. The Google AdSense ads look like ads – this is bad because users don’t like to click away from facebook because they really like being on facebook. Ami Vora says that half of facebook’s 50 million 30-day active users come back at least every other day. When they do come to the site they average 50 (!!!) page views per day. To average this amount, it means that many users are viewing hundreds of pages. This shows that the goal of these consumers is to stay on Facebook (not click on link-off ads).

2. The Google AdSense ads aren’t that relevant in a facebook app – this is because most facebook app pages don’t have enough text for the Google’s algorithm to figure out what type of ad might be good (unlike normal AdSense content network web sites where there is a lot more text on which to base the ad selection)

3. The Google AdSense ads don’t share a high % of their revenue with the app – this is actually a guess because Google in their arrogance / monopoly-status doesn’t tell the app how much Google earned from their app’s ads nor what % of Google’s earnings that the web site owner earns. This means Google can manipulate the revenue share with each app at will.

This should lead app developers (the energetic, business-oriented ones) to test the Facebook specific ad networks (e.g., Lookery, Appsaholic, Adonomics, Cubics, fbExchange,etc.) and other web-wide ad networks (e.g, AdBrite, Azoogle, etc.). In reality, most app developers don’t even bother running any ads because they don’t want to take the time to optimize this process (which requires testing multiple ad networks and/or changing out the ad types, ad offers and/or ad networks frequently to prevent ad blindness).

The Origin of the fbMonetize or fbCash Tag

What is really needed is a standardized fbMonetize (my name) or fbCash (facebook’s internal name) tag that would allow any developer to create iFrame ad areas inside of their facebook apps that would be managed by Facebook or via an open source group (or one of the major ad providers in a neutral way). The way this tag would work is that the app developer would provide the real estate in which ads could be shown and the ad networks would provide their best possible ad inventory and revenue share rates in a competitive bidding model to claim that ad real estate. The ad optimization algorithm would use the available ads and the detailed demographic data that facebook offers to pick the most lucrative ad for that particular spot in that particular app for that particular user. Currently, the best performing ads on facebook tend to be Cost Per Install ads for other apps because the user knows that these CPI ads will keep them inside of facebook (which is what they want).

Facebook’s Master Plan

Facebook’s plan is to take this type of concept of an fbMonetize or fbCash system (which will be of limited value inside of facebook because it is hard to get around the fact that facebook users don’t want to leave the site) and apply it to general web sites. Here is how it is likely to work:

1. Facebook (with Microsoft’s help) will offer a competitive solution to Google AdSense for non-facebook Web sites.

2. You can think of this service as an Open-Source AdSense solution where Google can provide ads into it (if they document what the web site owner will earn) but Google (and any other ad providers) will have to compete with ads that Microsoft can provide that are facebook-enhanced

3. The innovation here is that Microsoft’s ads will be able to pick up the user’s Facebook cookie (for the 50 million growing to 200 million users who already have a cookied-Facebook account).

4. This means advertisers in Microsoft’s AdCenter can offer a much higher CPC or CPM payment to the web publisher because they will know that the user viewing the web page is actually a facebook user that, for example, happens to be an 18 year old male with a birthday in 3 weeks who mentioned X-box on his profile page.

Clearly, this type of knowledge of who is viewing a destination web site will make all of the current Google AdSense content-site inventory more valuable but only when it is provided with ad inventory that offers higher rates for clicks from the ideal customers. Currently web developers only support Google AdSense because Google was the first to offer this service (and no other service has found a way to show ads that were more likely to be clicked on). However, this Facebook-Enhanced, Open-Source, AdSense solution that facebook/microsoft can now offer all web publishers will allow every existing Google AdSense web site user to start earning more revenue from ads that can be served in real-time through a facebook/microsoft system. In the cases where no facebook cookie exists and/or Microsoft offers no good inventory, then the algorithm can default to the normal Google AdSense offerings.

That is my informed guess as to what is coming from facebook.

To get a feel for how targeted Facebook ads could be, play around with the Cost Per Click flyers here:

http://www.facebook.com/flyers/create.php

 and click the targeting choices for men, 13-15, in the US, who mention Xbox in their profile and you will find out that there are 1,380 of these that can be reached.

Facebook’s Big Surprise

This is pretty amazing when you realize that every web site that offers ads could now be optimized with this type of data. No one really expected Facebook to make general web advertising better, but I think that is what will be announced. Strategically, this impacts a big chunk of Google’s non-searched based ad inventory and makes it even more problematic that Microsoft is not only keeping facebook a google-free zone but now is using facebook’s permission-based demographic targeting to help convince web publishers to move beyond just supporting the current Google AdSense solution.