Facebook is planning a new advertising system that will target ad delivery based on profile information added by Facebook users.
According to a Wall Street Journal report, the new system will “let marketers target users with ads based on the massive amounts of information people reveal on the site about themselves” with future development that is aiming to (in true Orwellian style…or should that Minority Report style) “predict what products and services users might be interested in even before they have specifically mentioned an area.”
Monetizing social networking platforms through advertising has been a hard task for a long time now. For example, although News Corp may be reaping revenues from MySpace today, most of those revenues comes indirectly, such as through the Google search deal. MySpace rates have previously been reported to be a small fraction of 1 cent CPM.
Facebook seems to be experiencing stronger advertising rates, with recent reports indicating that Facebook charges around $10 CPM. There is always scope to improve, and certainly the more upmarket demographic of Facebook should provide high paying advertising opportunities if and when the ad delivery is sorted out and well targeted.
There will be some who will question Facebook’s moves. Google has previously come under fire for delivering contextual advertising in Gmail based on the content of emails a user had received. Facebook is not only going to use user data and networking activities to deliver targeted advertising, they are going past that and trying to produce preemptive targeted advertising based on what they think you might want in the future.
If Facebook can create a system that accurately preempts user desires, it’s not unreasonable to consider that Facebook could easily become the next Adwords as well; not even Google can currently predict the future.









Obviously Facebook was going to use info for advertising. This is a qualitatively different approach to Google. And I see Facebook’s approach being the only other option in the future.
Now picture if Google owned Facebook, and how much strong marketing data they woudl have.
I didn’t realize this was news. Everyone knew this. Its why lookery got hosed and why they kept people like eggnetwork on the hook.
I think the targeting using user profiles isn’t totally new, but it’s the first I’ve heard on the preemptive side. Certainly the WSJ is reporting it as new.
Regarding the Orwellian style… Like if I’m in a Blackberry group, I’ll see ads about the iPhone?
I might be wrong but chances are these “predictions” will either be quite predictable (obviously) or terribly wrong. Either way, I think there’s more spin to it than truly smart technology at work for the advantage/benefits of advertisers and users.
BTW, what the hell happened to MySpace? Lately I read 50 stories about Facebook versus 1 about MySpace. They’re still a bigger network, you know that, right?
RBA
From what I can gather it’s a little more involved than that. The iPhone ads you are seeing are contextual (ie targeted because you are a phone user), this is more along the lines of you are on a Blackberry site and you get delivered advertising for Tahiti because they think you’ll be wanting to go on holiday.
Interesting point Ben. Put Google and Facebook together, and an acquisition of a mass retailer such as Amazon and you’d have a hell of an e-commerce and advertising system delivering perfectly targeted consumerism.
Hopefully Facebook incorporates a pay-per-click system, would be easier to control costs and achieve strong ROIs. But then again, when did an advertising system ever care about their advertisers when it came to larger profits for them (see Adwords).
Yahoo should buy Facebook… only chance to compete with Google.
Aydin.
http://www.bOKnow.com
Duncan, yeah mine was a bad example, but I certainly don’t see the relationship between using a Blackberry and going to Tahiti. Hell I’d go Tahiti anytime if I only could
As I see it, preemptive advertising works (should work) like this:
If today I’m writing about diapers, then in 2 years I’ll be seeing ads about preschools.
If I announce I’m engaged, soon I’ll start to see honeymoon destinations.
If I’m posting copyrighted content in my FB page, they’ll send me the addresses of some local attorneys at law.
If they notice I’m spending way too many hours in FB from work, I’ll see ads about Monster.com and CrunchBoard Jobs.
I don’t get it…
Mike post something for duncan while they in office. Do you talk to each other in office?
Why submit comments?
The MySpace rate is a small fraction of 1 cent CPM? What’s the source?
How can it be a fraction of a cent per CPM?
Let’s say 1/2 a cent per CPM. That means $100 buys you 20 million page views.
100 dollars * 100 cents per dollar * 2 CPM per cent = 20,000 CPM * 1000 views per CPM = 20,000,000 page views? I’ll take that.
Maybe you meant a fraction of a dollar.
adblock plus and/or adblock pro?
I’m not sure why more people don’t use these tools. It will be interesting to see what happens if/when they gain traction.
@50Cent: Your analysis is dead on. Tracking is sometimes more expensive than a fraction of one cent. My guess would be around 0,15 to 0,40 USD CPM. Assuming a click rate of 0,05% to 0,2%.
@Facebook CPMs: We (www.verwandt.de) tried to buy advertising from them for profiles based in Germany. Only offered their “flyer” (left site of profile). For 2,- USD CPM w/o any discounts. Even assuming a higher click rate of 0,3%, that would result in a CPC of 0,67 USD. I don´t think they can sell out their inventory at that price. A CPM of 10 USD might be a list price, but nobody will pay that.
Best, Sven
Warm it up Chris. I’m about to.
Facebook is a big yawn. Night.
My understanding from previous reports is the junk ad rates on MySpace is something like $0.00001 CPM give or take the odd zero. I’ll see if I can find the reference but I do distinctly remembering it as being less than 1c
Weird
I’m on the other side of the world to Michael, so no.
These targeted ads have been going on for some time on Facebook as well as other online applications, like Hotmail.
I have noticed a big difference with the ads displayed for the different genders of users. For example, female only motor insurance. I have also noticed aged based delivery of ads.
While at moment ad delivery is probably only based on basic user profile elements, I assume Facebook will be going further, like matching ads based on a user’s likes and dislikes (should they be specified). You might see an improved interface to encourage more users to fill out these details in the near future.
It would be very interesting (and an absolute monopoly) if google tries to acquire facebook. Going forward, will companies look at “user data” syndication given that privacy ceases to be an issue for most of the ad companies. I wont be surprised if that happened.
Ads are a part of orkut now, wonder if google are planning something there. (had to talk abt orkut, its big in our part of the world)
Hi Duncan,
First was Hotmail. I remember the times when people wonder if Microsoft reads your letters.
Then Gmail. They had a project to give more space if you allow to receive ads based on the context of your mail. For example: “Dear Mom, I did not write you because my typewriter was broken,” then will see a Smith Corona Ad to your right hand side. (I feel old) This was a failure as such; but look how this was the lead to “adsense for context.”
Now is Facebook. Based on your profile Facebook will target ads. Explained like this sounds like an abuse to our privacy. Sooner rather than latter, Facebook will begin to make money, although they will not shot at their own leg. How, we don’t know yet, but not in this way.
Mario Ruiz
@ http://www.oursheet.com
Duncan,
Geez – what a bargain! If it’s $0.00001 CPM, then I could buy 10 BILLION pageviews with a hundred bucks. Haha
Hahaa, you are funny, 50cent!
@Mario Ruiz : Why is contextual advertising based on data I’ve input more intrusive than that based on what site/blog/content I’m reading? It’s still all in the aggregate too…
As to the predictive side I see it being based on the aggregate again. For example if you like Blackberries and FB can tell that a majority of the users that used to list Blackberries now list iPhones they could market the iPhone to you. I’m amazed less sites do this sort of aggregate comparative advertising…
/Martin
The predictive ad system just seems like a profile-detail clustering system where new keywords get added to what the advertiser requested.
Interesting that this topic should come up. We experimented yesterday with Facebook flyers and didn’t see a dent in our pageviews (says Google Analytics). To make matters worse, there was no way to verify that the flyer even showed up (as our profiles are not in the networks we were targeting (to be in a college network you have to have a .edu email address apparently) we couldn’t check to see that it appeared)…and still no response from the Facebook people about this (sent them a message yesterday about the fact that I couldn’t verify that the “flyer” was up)… very dissapointed, and no chance we’ll ever advertise again on Facebook.
predictive advertising optimization based on user meta data is not new or revolutionary, odd that the wsj seems to think it is somehow. a couple of things that will be interesting to see going forward are facebook’s ad rate curve after implementation as well as the impact the results from the system have on their user base. if facebook does not have any unbiased collaborative data, it will make the system easy to “manipulate”.
It’s about farking time.
*knock knock* HELLO? MYSPACE?? *knock knock* ARE YOU THERE?
Finally. I’m still waiting for Myspace to do this as their ads are not targeted at all. I was wondering when either of them would catch on.
While I also agree with Info21dt above that, yes, it would be easy to manipulate, a majority of Facebook users have genuine information in their profiles so overall it would be better at targeting ads for that majority.
Is this even new? I had requested a Facebook Ratecard last winter, and the one they sent me allowed for “CPM Filters” in which you could target your banner ads to a specific school, or age group, or sex, or even “People who like the band U2″. Each filter was an additional $2CPM. Get real!
Sounds to me like it hasn’t done well at all, and their PR folks are trying to revive it? Honestly these filters were making the prices run into the $20CPM range. $20 for a social network? LOLOLOLOLOLOL
I think you (or perhaps the WSJ) may be taking the “predicting the future” thing a little to literally. That’s really what every form of targeted advertising tries to do: you place an ad for a product that you hope the user will be interested in. Targeting the ads means the end user is more likely than most to be interested in it, but it doesn’t mean you’ve successfully predicted the future.
If ads are going to appear on my screen I’d rather be “targeted” with ads that might be of interest to me.
The one who is going to come out on top in all of this is Microsoft. Check it out: Facebook is a “Social Networking” site, but in web2.0 land, one of these sites is the equivalent of the inside of your high school locker. The Social Networking term is pretty much abused and overused and no MySpace, Facebook, or other knows how to monetize any of the true Social Networking data they’ve captured. (just look at Y!, they’re sittin on piles and piles of data and they don’t even know it! They keep wasting their time with 360, etc.)
Now, when you target ads, you do this based on context, behavior, or ( and some see this is the holy grail of advertising: ) a person’s social network. This last way is not an easy thing to do, but guess who’s looking to get into the internet game they’ve missed out on for so many years and has an army of extremely intelligent people focused on this exact problem? Oh, yeah, that’d be Microsoft. Oh, and they just happen to have a deal with facebook through 2011. I wonder if something in that deal gives them any access to all the data that fb has captured? FB is the flavor of the month, but they could be gone in a while. There was a time when everyone was riding MySpace’s jock pretty hard. But Microsoft is going to stay paid and be serving you ads from every site in the future with the targeting they’ve perfected through the fb experiment. Ahhh, good times.
And I don’t think Y! should buy fb. They wouldn’t know what the hell to do with it. If they pulled their heads out they’d realize that with all their data they could crush everyone and be the leader in Social Networking in a very short time. Analysts, press, and blogs have told them they suck and for some reason they listened and internalized that. Which is great for everyone who wishes to defeat them as a business. If they get some self-esteem though they just might end up bitch slapping Google and FB and everyone’ll be rushing to ride Y!s jock. That’s the scenario I’d really like to see.
All you tech guys see things thru rose colored glasses. Do you not think people are sick of being slammed with advertising at every damn click they make. I know I am. My children are. My grandchildren even complain. If all these sites can do is gather a massive audience only for ad bombardment, c’mon, get a clue. Advertsing revenue may be great for the site owners, for a while, but then……..poof ! No matter what service they perform or thevalue they offer, the eyes can only take so much of this. Some of yu guys do a hell of a job writing code and understanding all the tec terms, but forget how the ” avg Joe” sees things.
Yup, I am older, but I ain’t dead.
mark zuckerberg seems a bit jaded
are there ads on facebook? i never noticed…
Ads, in general, piss me off. All intrusive ones piss me off – but even the easily ignored ones annoy me, because I have no interest in them. They’re useless to me, they’re irrelevant, and consequently they’re little more than background static as I consume information.
Ads, if I must see them, should be relevant to me. Useful. Communicate information to me that I might not have easily stumbled upon otherwise. As a 20-something guy, I’m sick of seeing ads for tampons, cars I can’t afford, TV shows I hate, and “ask your doctor about…” conditions no one under 60 has.
If an algorithm can look at my profile and friends on Facebook and figure out what sales, deals, and products I might actually be interested in – this is a win/win, as far as I’m concerned.
“If Facebook can create a system that accurately preempts user desires, it’s not unreasonable to consider that Facebook could easily become the next Adwords as well; not even Google can currently predict the future.”
Don’t you mean “predicts” users desires? They are not going to make a lot of money if they are preempting user desires.
pre·empt or pre-empt (prÄ“-Ä•mpt’)
v., -empt·ed or -empt·ed, -empt·ing or -empt·ing, -empts or -empts.
v.tr.
To appropriate, seize, or take for oneself before others. See synonyms at appropriate.
To take the place of; displace: A special news program preempted the scheduled shows.
To have precedence or predominance over: Discussion of the water shortage will preempt the other topics on this week’s agenda.
To gain possession of by prior right or opportunity, especially to settle on (public land) so as to obtain the right to buy before others.
v.intr. Games.
To make a preemptive bid in bridge.
Wow, that comment from Zucker is the most elaborate fake comment I think I’ve ever seen.
Nice job.
Wow. Facebook is working on a behavioral-targeting ad platform. How shocking! How revolutionary! I can’t believe no one’s ever thought of this business before!
Wait a minute, is this Facebook or Apple we’re talking about here? Sorry, I got caught up in the hype.
As an advertiser, I’m amazed at all the unchecked hype Facebook is getting. Even though it’s really just another site, I sort of understand the attention from a user’s perspective. I use it myself. But its financial potential and the hype being thrown about in that regard is laughably out of touch with reality. The comparisons to Google and Google’s revenue model are particularly ludicrous. Behaviorally-targeted ads, while a strong step up from contextually-targeted ads, are nothing new and still several rungs down the ladder from search in terms of performance. And contextual/behavioral models don’t scale nearly as well as search either. If they did, Yahoo’s revenues would be larger than Google’s. That’s why all the current valuations and rumored revenue potential are laughably out of touch with reality. Google already had the strongest platform in the strongest area. The fact that they were able to roll keyword-targeted ads out into contextually-targeted ads (across a MUCH broader reach than Facebook will be initially looking at) is just gravy.
Microsoft has had a nice, long peek under Facebook’s skirt by being their primary advertising network. They’ve seen the financial potential. They have by far the deepest pockets (and a current proclivity to spend it) than all the rumored Facebook suitors. And with the acquisition of Aquantive, they’re in a better position than anyone to monetize Facebook. But if the bottom-line value of Facebook really matched the hype, then Microsoft would be making an offer that would make all the previous offers (including the wildly speculative rumor of $6-billion) look like chump change. But the revenue (and potential for revenue growth) just doesn’t even come remotely close to the hype.
LinkedIn has been doing this for over 2 years already – the difference with them is, they actually have valuable information on people’s profiles…
$10/cpm!!?… Lets be real. Although the possibility lies to heavily target ad’s to users. $10/cpm is simply unreleastic for a network such as facebook and I’d be hard pressed to find any advertisers who can turn a postive ROI at that rate.
At 1.5 billion page views / day (according to zuckerberg at conference), that would equal $15m/day, equaling about $5 billion revenue per year. That would equate their market cap to be half the size of google. Maybe someday, but certainly not now.
@arrington, How did Lookery get hosed? We’re doing fine for now. We expect FB to do exactly what the WSJ reported today and are think we can complement them well. we’ll obviously need to build out specialized services that they don’t care about and we haven’t shown yet, but isn’t that always the case?
here’s a slightly diff interpretation of this same idea. we launched espin.com back in 2002, before it was kosher to go free-form social networking. so our interpretation of “closed” was that the profile building exercise was all done thru building blocks (radial buttons, drop downs, etc), which meant we had a pretty sophisticated database of info on our users. everything was tracked.
couldn’t sell the data via permission email as we never had the volume to justify premium cpms.
so we played it another way and created a totally targeted ad model within espin, where teens use their attention as a form of currency to access premium upgrades. if you’re a 14 yr old boy from peoria, il interested in wii, with a facebook page and into emo music — that’s going to influence what advertising teasers are shown to you when you hit the premium upgrade machine. advertisers tell us the demo/profile they want to buy, and those are the only users that get exposed to their teasers.
where we play it a little diff from facebook is that we don’t host the “canvas” page, or the ad unit for that matter. we simply channel our audience on a cpa basis to other people’s pages (to an ad, to a site, to a facebook canvas page — it doesn’t matter to us, just as long as the teen chooses it).
we’ve seen facebook’s beta efforts through their survey model. and it’s pretty cool. what we do isn’t that far different except that because we’re so much smaller (2MM uniques/month, yes we’re still a fairly closed site), we rally our audience and then ship them off.
Targeting advertising based on user profiles is not new.
Pingar has been providing this service to content providers for some time now. Matching online ads to demographic profiles allows advertisers to map their offline and online campaigns. It also allows content providers to provide premium content at no user-pays cost. Monetising premium content by targeting registered user profiles, whilst protecting their personal privacy, makes sense.
This is clearly something that Facebook has recognised. Others will too.
50 cents math seems bad – off by factor of ten $100 = 2M views based onnumbers not 20M
hey re rustys’s comments…”Behaviorally-targeted ads, while a strong step up from contextually-targeted ads, are nothing new and still several rungs down the ladder from search in terms of performance.” anybody have any good defensible statistical data on the validity of this comment? can you point me to it ? or just send it to me ? use my junk email smthngfishy@yahoo.com – thanks