Ad Firm Touts Results On Facebook Exchange
- by Mark Walsh @markfwal, September 13, 2012
Facebook in June announced plans to launch a real-time bidded (RTB) ad
marketplace allowing advertisers to reach users on the social network based on their browsing history. Through the Facebook Exchange (FBX), users visiting third-party sites can be shown ads related to
their travels around the Web when they go back on Facebook.
In short, it's a way for the company to expand advertising beyond its own walls via retargeting. They won't have to rely simply on interests listed in a user's Facebook profile and the pages they “Like” on the social network to target ads.
So Expedia, say, could drop a cookie on the browser of a visitor who researches a trip to Iceland but doesn't actually book travel. Expedia could then bid through a participating DSP (demand-side platform) to show that user an ad for a travel package to Iceland. Such an ad would presumably be more relevant than a more generic travel ad, leading ultimately to higher return on investment.
And higher ROI is what the Exchange is delivering, according to one Facebook ad partner, based on beta testing so far. Retargeting platform AdRoll says that after running 60 campaigns for clients including Room & Board, HootSuite and GoPro since June, ads sold through the Exchange have resulted in 16 times higher ROI on average than the standard Marketplace ads that run on the right side of Facebook pages.
AdRoll declined to provide further details, like eCPMs, click-through rates, cost-per-click in connection with campaigns run through FBX. The company was among the initial group of DSPs selected to trial the system along with AppNexus, DataXu, TellApart, MediaMath, TheTradeDesk, Triggit and Turn.
AdRoll on Thursday announced it will begin opening up access to Facebook inventory to all customers after gearing up its RTB infrastructure to begin serving campaigns more widely. “We've invested quite heavily in our RTB technology, which allows for precise bidding decisions based on users' anonymous browsing behavior,” said an AdRoll spokesperson.
Rocket Fuel and Nanigans are among other social media ad platforms announcing their support for FBX. Others added to the original group of Exchange parters include Criteo, Optimal, Xaxis and X+1, according to TechCrunch.
With FBX, we’re excited to help all of our customers improve the reach and performance of their campaigns through Facebook’s massive inventory supply,” said Adam Berke, president of AdRoll, in a statement. Facebook accounted for about 28% of all display ads served online last year, according to comScore.
In a research report earlier this month, JP Morgan analyst Doug Anmuth estimated that site retargeting generally carries a CPM of $2 compared to Facebook's U.S. CPM of about 45 cents. He estimated a 2%-3% gain in Facebook's U.S. ad revenue for each 1% of its impressions that are sold through FBX.
In addition to Facebook's top line, the Exchange is expected to benefit direct response marketers especially by opening the door to ads more focused on demand fulfillment than raising brand awareness. But more information about results and costs from early campaigns may be needed to help convince advertisers to get on board en masse when the Exchange gets beyond the test phase.
FBX is Great. However, comparing FBX (where you can bring your own targeting) to Facebook's regular right rail ads is apples and oranges. Of course bring your own targeting and segment or RTB works better than relying on Facbook's profile based targeting.
Exchange buyers often have their own data (site retargeting) or third-party data that has high signal intent. Regular Facebook ads have low signal of intent.
So, it's all well and good to be excited about FBX but the more important comparison is to the other exchange publishers. So far things look fairly good for some advertisers in that context based on our tests.
Hmmm. Sixteen times higher ROI. That is a BIG relative difference. Any chance of actually seeing the data to see the magnitude of difference? That is, is it a case of 0.03% went to 0.48% - from infinitesimally small to abysmally small?
This relative data was reported incorrectly in the article, and has since been corrected: http://www.mediapost.com/publications/article/182959/correction-adrolls-roi-on-facebook-exchange.html
@John, Triggit saw as much as 4x higher ROI across the board. You can read our full results here: http://blog.triggit.com/facebook-exchange/
We're also the first partner to offer treatment and control testing on FBX, and we found 18% - 30% conversion lift, and we're sure to be seeing results like these for the foreseeable future on FBX!
Working for B2B also, clearly a game changer: http://www.digiday.com/platforms/how-facebook-just-became-critical-for-b2b/
Thanks Christina, but the devil is in the detail and not the sales deck. The thing with ROI data is that there are two sides to the equation. The first is the 'return' - and that is what I would like to see further substantive quantification of. The second part of the equation is the 'investment' which in initial phases can be so easily manipulated by the vendor. Sorry for my cynicism, but I can 'prove' any medium can have any ROI it wants. I've even achieved an ROI of infinity - beat that! Yep, convinced a small TV market broadcaster to give away a package of TV spots to try to get the client to increase their budget enough to include TV. The client could measure the uptick at the till in order to determine whether it would be worth it on an ongoing basis. There was a healthy blip in sales which, when you divided it by zero, produced infinity times ROI. I was only interested in the single-digit percentage sales increase to show that 'it worked'.
@John Agreed, and well said. Every advertiser attributes conversions somewhat differently, and there's no one size fits all ROI calculation. Given the early data, we tried to provide a sense for how the first round of campaigns have performed based on widely used attribution modeling. As is evident from the range of data points coming out of the beta, we're all still early in understanding the ins and outs of this exciting new inventory source. However, I think the underlying point is that initial results are very encouraging, no matter how you slice/calculate it.
Thanks Adam. I'll settle for 'very encouraging'.
Kevin Lee from Didit points out that Exchange buyers often have their own data (site retargeting) or third-party data that has high signal intent.
My question is, do they (as I suspect these early FBX results may) have the same data - as that of the 'integration (of) Search and Display', I wonder?
Data (that means) when a search engine user enters 'home mortgage' in a search, your ad could appear next to the search results shown to that searcher or, when that searcher is next found, on Facebook?
http://seekingalpha.com/instablog/36191-lookingconfident/932031-an-rtb-global-open-marketplace-handling-up-to-100-billion-events-per-day
@Ross The idea behind FBX is that it gives the ability for advertisers to target users on Facebook based on any data they have. This includes first party data (site retargeting) and 3rd party data (which could include search activity or behavioral data from other sites around the web.) Basically, any data that advertisers could historically target on through a DSP is now available for targeting on Facebook. Hope that helps to answer your question.
Thanks for your great reply Adam Berke from Adroll.
I was leaning more to the 'evolution' of what I'm now terming as being, "Search targeting". - I've explained it more in my blog-post just now. - Just thoughts.
http://seekingalpha.com/instablog/36191-lookingconfident/1074411-x01-x01-x01-x01seo-s-say-goodbye-keyword-referrer-data-making-way-for-a-one-open-marketplace