Yesterday, Facebook took its first baby steps in developing its own true mobile ad network when it began testing its new mobile ad platform. The key take away on the new platform is that Facebook is looking for ways to extend its ad capabilities outside of Facebook, and to find ways to deliver ads without doing harm to its own Facebook user experience. The core idea is to use Facebook data - that is, to use the information Facebook has in hand about its users, to allow advertisers to offer meaningful ads to Facebook users no matter where online they may be at any given moment or what mobile apps they may happen to be running at any given moment.
The forerunner of this approach was Facebook's initial efforts to pace ads on Zynga.com. The tight relationship between Zynga and Facebook helped accomplish this, but the new mobile ad platform can't rely on such tight relationships. For it to work it needs to do so across an extremely wide ecosystem. Further, it needs to do so in ways that are unobtrusive and in ways that don't leave Facebook users feeling somehow "violated" in terms of the information they place on Facebook.
Image via Shutterstock
What does this mean exactly?
Facebook will work with advertisers, marketers, publishers, ad networks and exchanges and ultimately anyone else who is interested in delivering targeted mobile ads to utilize Facebook user data to determine where such ads should appear. What data will Facebook share? The short answer is any data Facebook already uses to deliver ads on actual Facebook pages. This includes a subscriber's age and gender, user "likes," user location, and other data Facebook has permission to use.
In particular, Facebook will also make available any associations Facebook is able to make between a given advertiser's products and subscriber friends who have used its products. The idea of course is that friends have similar interests, so making the leap of faith that a friend's use of a product (which in addition may have gotten a "like" from that friend) may also be something a specific subscriber may very well be interested in.
This can be a powerful means to target users on an enormous scale - given Facebook's enormous user base the company should be able to make huge (or at least non-trivial) numbers of such associations and then be able to refine targets through other Facebook data. Certainly advertisers can appreciate the potential value here.
How it Will Work
In theory, Facebook will collect fees from advertisers and will allow advertisers to target subscribers with advertisements based on the Facebook data noted above. Initially these ads will take the form of traditional ad banners and other interstitial ads. So far, Facebook has not made public any of the advertisers or marketers participating in the initial test, but clearly there will be major players taking part in the grand experiment.
Here is the key difference that Facebook allows: The ads will use specific Facebook data but will appear outside of Facebook. Ads will be placed inside of other mobile applications and mobile websites but not on Facebook pages. This is key to "extending" Facebook's reach outside of Facebook. Keeping in mind that most - if not all - Facebook subscribers will have multiple (and potentially numerous) destinations in their profiles outside of Facebook itself, what this means is that Facebook can easily open up far more access than its own headcount might suggest. Those 900 million users can turn into, say, three billion destinations for advertisers to target if we assume a mere three external destinations per user outside of Facebook.
The Facebook ad network lets advertisers and marketers place bids on Facebook's ability to reach any given demographic - for example, single women looking to buy perfume of a certain type. If a Facebook user (who is identified through anonymous Facebook IDs synced with various ad exchanges) visits an app or site where the ad exchanges have placements and there is a suitable match, the exchanges in turn will send Facebook the anonymous user IDs and then checks to see if any advertiser has placed a relevant bid.
Whenever there is a match - and given the size of the Facebook audience the odds are good that suitable matches will be found (this is key), Facebook will pay the ad exchange some percentage of the bid amount and will then deliver the ad to the user. It's a very interesting mobile approach that completely factors the actual Facebook site out of the advertising equation, while ensuring that the Facebook sire remains core to Facebook's ability to deliver exactly what the advertiser needs.
Back to the Future: Traditional Advertising Methodology
In earlier analysis of Facebook and its advertising plans we have suggested that one major way Facebook can monetize the data it has on hand for its users is to take that information and apply it to TV advertising. The platform described above begins to take Facebook in exactly this direction.
The initial approach the new platform will take is to allow advertisers to place banner and interstitial ads inside of mobile applications and mobile websites. There are, of course, thousands of relevant mobile websites. The mobile apps can be either Apple iOS or Android apps, which opens up a vast area of advertising possibilities - the current count on mobile apps within these two platforms is 700,000 iOS apps and 600,000 Android apps - that's 1.3 million mobile apps overall!.
Ads will behave traditionally in this first effort. They might take users to other app store pages where users can download apps or free sample apps, or to mobile websites that might feature whatever services or products advertisers have targeted users for. One interesting thing to note here is that there is nothing inherently social going on. The new Facebook ad exchange service ties directly back to traditional advertising methodology.
Facebook can ultimately transform itself a huge meta-exchange of sorts. What this means is that the Facebook site serves as the data source while external outlets become the means to generating Facebook revenue! This can be huge for Facebook - it would allow Facebook to retain a neutral presence on its website - and would allow it to remain "Facebook without ads" - something that most Facebook subscribers would claim to want. And if done right, the underlying data gathering won't strike users as either obtrusive or otherwise invasive.
It isn't going to be difficult to make the leap to other platforms. Facebook-based TV advertising is a logical eventual step. The value here is that it takes Facebook data and exports it to other major ecosystems - that is the real key, some of us believe, to how Facebook will earn back its original IPO valuation. If the current test succeeds it will be time to get back on the Facebook stock bandwagon. In fact, it may make Facebook's stock look like a good buy today.