Posts Tagged: tracking

Is Facebook Atlas the Google AdSense Killer?

Alex Houg will be speaking at Pubcon Las Vegas on Thursday, October 9th, where he’ll discuss Facebook Atlas and other topics in-depth. Check out the panel “Real World Results” at 12:40pm-1:35pm in Salon E.

Not so fast.

Facebook announced this today, creating broad speculation by non-marketers that this is the event years in the making.  Finally, a response to Google’s AdSense. The Atlas ad serving platform, bought from Microsoft, which in turn, was part of the Razorfish (agency) acquisition, is just an ad server.

2014-10-09 03_25_10-Is Facebook Atlas the Google AdSense Killer_09302014_v2 - Microsoft Word

Warning: geek talk ahead, but worth it for practical-minded, non-geek marketers.

The talking point of “people-based marketing” is strong. It means that Facebook is targeting based on what we know about people via the Facebook login, as opposed to only cookies.  While cookies are faulty across mobile, which is 60% of Facebook’s traffic, a true marketing solution must accommodate tracking users by cookie (website pixels), email, and native userids.

2014-10-09 03_25_46-Is Facebook Atlas the Google AdSense Killer_09302014_v2 - Microsoft Word

To show ads triggered by remarketing and general social targeting is certainly interesting.  Facebook’s existing data platform will eventually squish 3rd party cookie collectors, DSPs, 3rd party ad servers, and most independent data providers.  There will certainly be a hold-out audience of folks who still buy on insertion orders, for negotiated impression-based rate card deals.

The discussion of tying to offline sales is not Atlas-specific.  Facebook already has conversion tracking and Datalogix integration, the latter of which is only to big brands with a large retail footprint.  Facebook’s cross-device tracking is perfect for mobile and only they have the scale necessary to tie users. DSPs are dead, as well as many in the inefficient middle.

If this sounds like technical ad geek stuff, this is what you need to know:

  • If you’re a small business, you don’t need to know anything about ad servers.You’re not selling your inventory for money, and even if you were, you don’t have enough to make it worth your time. Your customers are worth more than a few pennies of selling banner ads. Don’t run ads on your own properties and don’t buy ads except directly on Google and Facebook.
  • If you’re a marketer of any type, you need to first focus on understanding marketing automationbefore you worry about negotiating traffic to buy or sell. Unless you must buy or sell banner ads, focus on buying on native platforms (directly via Facebook and Google).
  • If you are a major publisher (over 10k uniques a day), this is your answer. You probably won’t be kicking out your ad sales team, but you can look at replacing stuff from other vendors. The cookie-only guys don’t have an answer here, especially on cross-device tracking.
  • If you’re a technology player, especially an ad serving player, time to update your resume. You knew it was a matter of time before the money got connected to the technology.
  • If you’re a student, learn all you can about the data you can collect from all free platforms(Google, Facebook, LinkedIn, twitter, etc).  These guys let consumers play for free in exchange for data that powers ads. So you’ll need to help businesses bridge this gap, to use data to help drive smart marketing and sales. This trend is not going away.

Here is their product tour.

Facebook has their javascript tracking on over a million sites, last time we heard a couple years. Who knows what the number is now. With the roll-out of Altas, we can expect this number to jump.

Javascript pixels allow Facebook to collect an immense amount of information about users on a site.  And when they authenticate to Facebook, their system can then tie that web user to a Facebook userid, allowing Facebook to track the user into mobile and cross-platform.

Nobody disputes that Facebook has the widest footprint of any network, cookie-based or not. What we don’t know is how much of this web-collected and natively-collected data can be used to help us as marketers. A lot, I’d imagine.

What do you think about this?