Now, as Facebook has begun selling ads in a different, much more lucrative way that others have been doing for years, it feels like Facebook’s tactic has put the company years behind schedule.

How Facebook has sold ads for most of its history: Advertisers tick off a bunch of boxes on the type of people they’d like to reach and then Facebook shows their ads to these people.

That’s the “old-fashioned way,” because that’s basically how advertisers have been buying TV ads for decades. Advertisers have a group of people to whom they would like to market their products, and they look for the TV shows that have that audience.

In other words, they look for inventory that is targeted to an audience based on data a publisher provides about its audience.

For the past five years or so, Facebook has sought to improve on this old-fashioned model by giving advertisers more detail about the type of people they can market to.

Instead of just knowing where those people are located, their gender, and their age, Facebook can tell advertisers where the people viewing ad inventory work, their marital status, and what their “interests” are.

All this extra data was supposed to be a gold mine for Facebook, and Facebook built up a huge ad sales apparatus to sell ads targeted with it.  

Eventually, Facebook’s ad business grew to annual revenues of just under $5 billion per year.

That sounds pretty big but …

It turns out this whole tactic may have been a big waste of everyone’s time.

That’s because in recent weeks, Facebook has begun selling ads in a new way that makes its massive inventory much more valuable—three times more valuable, according to one company buying the inventory and reselling it.

This new method is called re-targeting. It has been used by ad-sellers outside of Facebook for years now. 

Facebook, however, only began selling re-targeted ads this summer, when it opened something called the Facebook Exchange, better known as FBX in the industry. 

FBX is what it sounds like: an exchange. Facebook has selected a dozen or so companies that will buy Facebook ad inventory and sell it to marketers using re-targeting.

How re-targeting works: You visit Warby Parker, the online glasses seller. You look at a pair of glasses you might like to buy. You decide not to buy them right then. You leave the Warby Parker website. Later, on other Websites you see ads with the pair of glasses you liked.  

You see those ads because when you visited warbyparker.com, your browser downloaded a tiny piece of software, called a “cookie,” that told the ad servers on sites using re-targeting that you had previously gone to warbyparker and looked at a certain pair of glasses.

Ads that are “re-targeted” in this way are clicked on a lot, and it’s pretty obvious why. Unlike most ads in banners on the Internet, re-targeted ads are ones that you may actually want to see because they are based on your demonstrated interest in a product.

Because they are clicked on so much—and because those clicks so often lead to sales, re-targeted ads are valuable, and publishers are able to charge advertisers steeper rates for them.  That’s good news for Facebook.

How good?

Zach Coelius, CEO of Triggit, one of the ad-reselling companies Facebook has invited onto FBX, says that return on investment for advertisers buying through FBX is so good, that if all of Facebook’s ad inventory were sold with re-targeting, instead of user data targeting, Facebook would be able to charge 3X the price it charges for ads right now.

What’s truly remarkable is that inventory sold through FBX re-targeting uses ZERO Facebook profile data, and yet it is much more valuable. This has to make you shake your head about Facebook’s strategy for the past few years, and we’ll get to that in a second.

Coelius has a stake in seeing FBX do well, and having business news outlets write stories about how Triggit is getting great ROI for its clients. So you have to take his claims with a grain of salt.

That said, his claims make intuitive sense. 

The most valuable inventory for re-targeting until now has been Yahoo Mail, because:

  • It has huge scale.
  • It’s engaging enough that you’d only want to click on an ad to leave if you really wanted to leave.
  • The people who use it tend to leave it open as a tab in their browser all day.

In all three ways, Facebook.com is very similar to Yahoo Mail.

So, when Coelius says that 18 months from now, most of Facebook’s ad inventory will be sold through re-targeting, and that rates will have gone up by a couple multiples, we find him to be credible enough. 

On a basic level, of course ads targeted to me based on my demonstrated commercial intent will be more valuable than ads based on what I put in my Facebook profile.

OK, I “like” the British Open golf tournament, and that probably means aspirational brands could do well to market to me, but that seems like a lot of guesswork compared with showing me an ad for a pair of glasses I almost bought two days ago.

Because it knows what marketers drooling-ly call “intent,” re-targeting feels a lot closer to search than Facebook’s profile data targeting does, and Google has shown us how great an online ad business can be when it is built around intent and massive scale.

The frustrating thing for Facebook shareholders, especially the ones who bought in on the IPO, has to be this: Facebook may be new to re-targeting, but re-targeting isn’t all that new.

As noted above, Yahoo has been doing it, and doing it well for years.

It’s ironic—and a little tragic—that Facebook spent so much time chasing a “new and improved” form of advertising, when an existing online tool works so well.

The good news is that re-targeting should be a growth engine for Facebook, which desperately needs one, even if we don’t think (ahem!) FBX meaningfully boosted Facebook revenues in Q3.