You could make the same argument for why advertising won't work for TV, newspapers, and magazines.
Google's model is nearly optimal in many cases, but I believe many advertisers still spend the majority of their money outside of Google. The reason is that there are two kinds of advertising, roughly: meeting demand, and creating demand. Google is good for the former ("I'm looking for X"), and TV has been great for the latter ("want to be important? You need this expensive car").
Right on about Need vs. Awareness .... Think of Coca-Cola or McDonald's, or even Budweiser. There are many large companies that use TV/Internet to keep their brand strong. Everyone knows about these brands, there is no direct "need" for any of their products (one could argue against this but I digress), but they need to keep advertising so when you are hungry, or thirsty, or sober you turn to them first.
That´s right, I think there´s also a hybrid, probably like banner advertising, google adwords, amazon recommendations and gmail, which is showing you ads for things you might be interested in but not necessarily actively seeking.
When you see an ad for coke on the street there is just some probability you are thirsty or will remember coke when you become thirsty, with adwords and other kinds of contextual advertising there is also a probability you might be interested in what the advertiser is presenting you, so they might just hit a need or making you aware of it, and not really creating it as in creating demand.
Great point. I was just watching some Saturday Night Live clips on NBC's web site. I was surprised that so many commercials one's forced to watch in their video area do not fit the mood of the site at all. Even on the Internet, when the commercials could be targeted better, they are not. This shows that traditional advertising networks that are transitioning to the Internet are still following the old approach, and it takes a new company to react to solve new problems and cater to Internet users rather than thinking of the Internet as a cheaper distribution source for traditional methods. Old media wants a "presence" on the Internet, new media lives and breathes the Internet, and is made up of regular people who are the stars to their social circle.
Matt's essay is actually consistent with what you said if you read all of it except the title. It seems like it should have been titled "Why Facebook's Ad Revenue Will Stay Less Than Google's."
We don't know that either. For example, the old-world analogues of Google are the classified ad section and the yellow pages. Both of them cater to people who know what they're looking for, and so have excellent focus. Yet I'd imagine that many, many times more money gets spent on TV commercials than on classified ads. It seems like "creating demand" is a much bigger market than "meeting demand", and if FaceBook can succeed at the former, it could be many times bigger than Google.
>It seems like "creating demand" is a much bigger market than "meeting demand"
I think that it's just easier to spend money on creating demand - Maroon's whole point about Google being the theoretical limit for targeting is that it's a hard task. Most people spend much more free time consuming media than discovering things, and it's the search part of discovery that is applicable for targeting.
Hmm, I wonder if you TiVo or the cable companies could mine your channel-surfing behavior to improve ad targeting (but that would require customized TV feeds instead of broadcast, yada, yada, yada). That would be interesting to identify other "discovery" activities that people do.
I'm not trying to defend his essay, merely stating that it argues something different from what's in the title.
That said, I'd caution against using your imagination to assess ad spending. Ask the nearest small businessperson what they spend on yellow-pages listing.
I don't disagree with you, especially in that television advertising isn't going anywhere (though DVR is clearly reducing value) but that doesn't refute the fact that Facebook won't make as much money off of online advertising as Google.
Supposedly a picture is worth a thousand words. I wonder how many pictures a video is worth. Or how many 468x60 internet banners a full blown page in a magazine is.
I think the big disadvantage to internet advertising is it feels so cheap, even when it isn't. When you see an ad on ESPN, you know somebody spent a lot of money to put it there, and that lends credibility. When you see an ad on espn.com, you don't get that sense even if the company happened to spend more online than the guy on television.
I would love to know television's effective CPM rate versus Google's
Yep, that's because Matt's really arguing that Facebook can't make money using the same model as Google, even though the title of his post implies he'll be talking about advertising in general.
He certainly didn't explain how "advertising works," only how search advertising on Google works.
Even more towards the point of the essay, users of Facebook are not looking for products and services. Facebook lets people keep tabs on each other and maintain more acquaintances but almost no one goes to Facebook to locate goods or services.
The problems is not only that Facebook lacks the ability to pinpoint precisely what its users want in the manner that Google can, it's that Facebook is the wrong environment to sell specific things to people. Again: people use social sites to keep tabs on people, not to buy stuff or look for specific items or services. The only ads worth showing on Facebook are those that emphasize brand loyalty, and frankly, the money is better spent on television ads.
He goes on to suggest that MySpace are the ones capitalizing on something like "demand creation", by realizing that people go there to waste time and therefore MySpace is doing a lot in terms of entertainment properties. Matt doesn't use the same terminology you're using, but I think he's getting at a similar idea.
He also says that creating demand is less profitable, so if the best you can do with your ad inventory is to create demand, you'd better have a much lower cost structure to accommodate that. I can't see into their wallet, but I'd imagine Facebook is burning through money pretty quickly.
I have to agree with the point Matt makes, that advertising will continue to work better on Google than on Facebook. We mustn't forget, however, that advertising is a construct of the last two centuries and especially the period 1950-1990, and may already have passed its prime as a way of generating customers.
Moving forward, the value of communities may overtake the value of marketing and of brand - and think here of baby boomer 'retirement' villages as much as MySpace or Facebook. Advertising will continue to work better on Google; but advertising as a means to generate business may be superceded.
I'll admit it's certainly based on a little conjecture. The post-WWII 'Product Era' supported by marketing evolved into the Brand or 'Positioning Era'; one of business frameworks I use for work (I'm a Shirlaws Business Coach) tracks revenue through a business, starting with Product, then Positioning, then Distribution (ie, of product to market).
But this does fit into emerging trends, both online (Facebook) and off ('baby boomer community' returns 5.3 million pages in Google, but no ads...)
"So if social networks are such a great experience, why aren’t people paying for them? Well, the networks let people sign up for free, because they want to have everyone on their network. But then they’ve immediately set the market value of the network at $0! Also, social networks are only a good experience because interacting with people is a good experience! If social networks vanished, people would still have friends, but if your friends vanished, you’d have no use for the social network. Whereas if computers and TV went away, people would still play games (board games? sports?), act, sing and dance."
I agree with this somewhat. Traditionally businesses have been dependent on advertising or PR in order to promote their product, but increasingly there is value in inflitrating certain networks (e.g. early adopter crowd) so as to spread "virally". You could call this a new type of business promotion e.g. community adoption or maybe its just a new subset of PR, in which case, nothing much has changed.
> "But they’ll never know what you want right now, and Google always does."
Matt, this appears to be your killer point, that search allows advertisers to know what you want right now. The limitation with search however is that there are things I am very likely to want that I don't search for.
Let me use an example. Often when people are about to go on holiday they announce it in Facebook by writing on someone's wall or updating status etc. Now if I said I was going to Cancun for example would't it be great for advertisers of tshirts, shorts and holiday insurance to advertise to me? Only Facebook can do this, Google cannot (I would have to do three separate searches, with FB I do nothing).
I think this goes beyond a friend recommendation because it links exactly to my plans. Its not only targeted but it shows what I want tomorrow. I think overall this is more targeted than google thanks to the demographic targeting placed on top which Google does not have. In Facebook, I could use the persons place of employment and age to work out their salary and offer two different holiday insurance schemes to the wealthy and the less fortunate. If I type in Holiday insurance in Google, advertisers cannot filter like that. And would think this likely to return a better ROI than Google search (as long as people pay attention to the ads! - the assumption being when ads are relevant people actually do).
So what I am saying is that there is a whole lot of data for Facebook to play with, don't rule them out yet. It will be very hard for them to pull this off but someone has to challenge the google monopoly and who are better placed than them.
I don't understand your logic. You type Cancun into Facebook and it hits you with "tshirts, shorts and holiday insurance". Why couldn't Google just do the same when you Googled for Cancun? That's not what Google does hit you with, but that's only because the market has decided that those ad slots should belong to travel agencies, etc. If ads ever took off on Facebook, it would presumably decide the same exact thing there.
I don't know about you, but like 99.9% of the population, when I travel to a new place I type it in Google first. I might type it into Facebook if I want to see if any of my friends will be there or have been there. But generally I'm Googling for what to do. It's gotten me through many new cities.
Also, you're giving Facebook too much credit for being able to parse natural language if you think this can scale for them. They might be able to target ads around specific keywords, but even that is going to be tough.
> "You type Cancun into Facebook and it hits you with "tshirts, shorts and holiday insurance". Why couldn't Google just do the same when you Googled for Cancun?"
That's the thing...with Facebook you don't have to search in order to be presented with relevent advertising. FB could parse natural language from the communications you are having with friends to display ads relevent to what you want now. Using my example above if i posted on my friends wall saying I am going to cancun then I would see the ads described above. However, that doesn't mean I woulnd't search in Google.
I agree, natural language parsing is very difficult but if Facebook could do that they may be on to a winner.
Even if the problem of generalized natural language understanding were solvable - which I doubt -, my bet would be on Peter Norvig and the Google team solving it way before the Facebook guys.
The fundamental problem is that natural language processing (NLP) isn't one problem to be solved. There will be many solutions to many particular problems of NLP. Many, many years later we're still waiting for one solution to energy (steam, or coal, or nuclear...?) or one solution to flight (gliders, or jets, or rockets...?). Human intelligence is a swiss army knife where the tools developed millions of years ago (opposable thumbs) or evolved withing ten years (keyword-based search). Why should there be a general solution?
Google has taken a sizeable chunk. But what's interesting to me is search has become all-consuming for them. Many solutions are perfectly fine for search, but that perspective (aggregating large noisy datasets) won't help much in other areas (e.g., individual voice recognition). If anything it will be a big waste of time trying to shoehorn every NLP problem into that search-based rubric. That leaves a lot of room for little guys cranking through data to produce new, and unique, NLP tools (e.g., Dragon Naturally Speaking). More to your point: I have yet to see a scientific hire by Facebook away from Google. It's been all business and marketing from what I've seen in the popular press.
Why indeed. However, finding one was very much on the agenda for GOFAI, and remains on the agenda for various AGI efforts, e.g. by people like Pei Wang, Ben Goertzel, and Eliezer Yudkowsky (who comments here, as I've observed).
All of the LBS (Location Based Services) haven't figured this out yet, but Location does NOT imply intent. Walking past a pizza store does not mean I'm hungry for pizza.
The question for Facebook is whether any data that they possess does a reasonable job of implying intent. It doesn't have to do as well as search, but it has to do well enough.
The question boils down to "what do you do on facebook that implies intent". The status may be one of the only things.
Knowing someone's intent is a sufficient, but not necessary condition for effective advertising. I may have no intention to attend a concert this weekend, but if I saw an ad for a nearby concert by an artist I like I would be likely to click on it. For this kind of advertising, Facebook can be much more effective than Google.
How? Everyone says this, but it just isn't happening because it isn't true. I mean, how does Facebook know that concert ads are better for you than t-shirt ads? They just magically parse your wall comments and this occurs?
My argument wasn't that knowing someone's intent is necessary. It's just that it focuses the advertising much better and as such demands far higher CPMs. Facebook can guess all they want, and if they do a perfect job (far better than technology allows currently) they'll guess as accurately as Google does now.
Facebook doesn't "know" that concert ads are better for me (although it can tell from my profile what music I like, where I live, and maybe even what concerts my friends attend, etc), but it can try different ads and see what works because of the large number of page views in a typical user session. In Google, you search, get the results, then leave the google.com. In Facebook, you can easily navigate to dozens of pages before logging off. This gives Facebook the opportunity to try different ads and see what works and what doesn't.
I agree with your main point, but the bottom line is that even if Facebook's CPMs are lower than Google, it doesn't mean that Facebook can't make a ton of money from advertising or that advertisers can't get good ROI on Facebook. You don't need to be Google to be a successful advertising platform. Also, don't forget that Facebook is still young and I'm sure many advertisers haven't yet adopted it. I have a friend who does internet marketing and he says that, at least in his niche, Facebook CPMs have been increasing by 10% a week.
It's true that a lot of LBS advertising models don't center around intent, but advertisers will still pay a premium for the additional context and relevance of reaching people near the potential point-of-sale. And what about LBS that employ Mobile Search?...
Then don't read them. Complaining about what other people write is the height of vanity.
I don't see how a top 10 traffic site could possibly be overlooked by marketers. If it were a value, they'd shift there until it was not, and that isn't happening at all. The data shows that you are incorrect.
I didn't read it, just briefly skimmed, my comment is related to seeing the reoccurring fb rants on YC.News.
In case you've missed it there is a lot of advertising happening on fb, some major brands include Blockbuster/BMW/Virgin America.
Where I see the value on fb is search ads for smaller advertisers, the cost-per-click on some prominent keywords are so much less than google, if the fb demo is your target, this can be a very cost effective channel for reaching them, plus this is where they are spending time.
Facebook could succeed if they figure out the "word of mouth marketing strategy online. Think about it; when you buy something offline, or even online most people ask their friends and family what they bought, and from whom ... and then make their decision based on recommendations.
When I pick up new web design work, the #1 vehicle is word of mouth from my already serviced (and hopefully happy) clients recommending me to their friends.
If Facebook can use the "Twitter" style of outsourcing questions to your friends, they can tackle this problem of 'focus'. I could ask my FB friends what type of music player they all use, and my responses could have targeted ads based on what they say.
FB needs to be careful though becayse as you said, people login to FB to "hang out" and don't think "buy buy buy" first like they do with Google.
1. Facebook doesn't need to be bigger than Google to be successful. All they need to be is "profitable". Why do you think they need to be an order of magnitude bigger?
2. You're missing the distinction between brand advertising and action oriented advertising. When Budweiser advertises, their goal is to create an association in your brain that links them to something cool you like so that when you're in the bar you choose their product rather than another brand. How many people do you know search for "beer" on Google to choose which brand to buy?
3. Facebook does have aspects that allows them to get that order of magnitude of views you mentioned. For example, if I "fan" a Facebook Page, many of my friends will see that in their news feed. Probably not all 250 of them, but many.
Online advertising is heavily weighted towards direct sales. I haven't used facebook that much but from the ads I've noticed, I don't see much branding happening - they're all looking to get clicks and convert sales.
Brand advertising on the net is all about clickthroughs too, just like any other ad. That's why you never see a banner with no link. The brand advertising is Budweiser's website, not their tiny banner. Since clickthroughs are whats relevant, the same laws apply.
Eh, the click is there so they can measure it and get more transparency than a billboard. But even if you don't click on it, it's about as effective as a billboard - not useful for a specific sale but one more incremental bit of brand awareness.
That's wrong because it isn't the tiny banner ad that is the billboard, it's the website you get to when you click it. Budweiser doesn't sell beer on their home page, they advertise it.
But I think reality is slowly setting in. Because what we’re finding is that when people want something specific, they go to Google. When they want to kill time or just communicate with friends, they go to Facebook. And that isn’t changing.
Let's say we split the market into two parts: Consumer A already knows he wants a particular product -- advertising is just a way to get him to buy it from you. Consumer B doesn't know what he wants -- he doesn't want anything just now. Advertising is a way to change this.
Isn't a list of Consumer B's interests, hobbies, friends, their interests, their hobbies, etc., a fantastic resource for targeting him? And isn't he an incredibly valuable customer? Apple didn't succeed by telling existing Mac users to buy more Macs (though they do!); it succeeded by getting people to switch. If Apple could target people whose status is "SP3 crippled my machine," or "Zune broken" or whatever, I'm sure they'll pay more for those users than they will pay for "Buy new Macbook pro" to link to Apple ads for Macbook pros as well as the Apple website selling Macbook pros.
Google can do a much better job with consumer B than Facebook. Facebook has to guess based on his wall posts, listed interests, etc. Given the sad state of natural language parsing, how well do you think this will work? At best they'll look at the bands I like and recommend similar ones. If they're lucky, they'll end up as good at that as Pandora. Or Netflix for movies.
Google knows your interests based on what sites you visit. Perhaps you visit an auto blog, which has Google Ads. Then you visit MySpace, which has Google Ads. Google shows you auto ads on MySpace. This sort of thing is already commonplace with ad networks.
Google has your entire search history. This is a far more effective way to determine your interests than Facebook. Facebook can't get beyond what bands and books and movies you like right now.
Something helpful to remember is the shift from broadcast to cable TV as a true innovation in advertising. Where the former meant huge audiences (this year's Superbowl still trailed the 100 million that watched the final episode of MASH in 1983), it also meant significantly fewer targeted ads, except for some time periods (detergents begat soap operas) and events (beer begat Bud Bowl). Cable allowed channels to become really geared to demographics and thus ads. It's an example where the long tail of content is quite valuable. Take Scripps Networks (Food, HGTV, DIY) - that's quite a sizable chunk of what folks consume on a weekly basis.
To many, the internet advertising model takes the cable approach and amplifies the number of channels. The difference with Google is they understood how clicks were the next generation game changer (and well ahead of their time). Yahoo had always focused on aggregating content. They were following a shotgun path that had proven very valuable (cable TV) when clicks were finally used to create a sniper's rifle for aiming ads.
What about trust as a factor in engagement? According to this survey http://tinyurl.com/3jcxjj recommendations from friends are trusted more than any other source of information. This is what people are looking to exploit with the social graph. Facebook beacon was an early clumsy attempt. If you can persuade people to pass on your message to friends you will significantly increase the response rate.
An example of this in action is at the end of this article on Duncan Watts http://tinyurl.com/3pgps7
"
Typically, people ignore this "share with your friends" pitch. But Watts and Peretti included technology called ForwardTrack, which displays the route the ad travels once you've forwarded it. This turned ad forwarding into a piece of social cartography. People would pass the ad specifically to those friends most likely to keep it moving. It became a Facebook-like contest to sign up the most friends.
"
Perhaps the key to making social network ads work is knowing where people are. AdWords has made marketing many, many internet only niche businesses possible. Privacy issues aside, Facebook could allow local bricks and mortar businesses to reach people geographically nearby via the internet.
While Facebook is a great way to keep in touch with friends who don't live nearby, I'm also friends with people that live in the same town and I see all the time - to see photos, arrange events, etc. You can join a group for the city you live in.
So, Facebook knows where people are and that some proportion of their friends live there too. Now there is new market for offline businesses - garages, restaurants, bars, shops - to reach local people on the web.
There are even Facebook groups for people that work in the same company. I'm sure the pizza place across the road from one of their buildings would like to advertise their promotion to everyone in that group.
I think Matt has brought up a very good point. He is right for the most part. Facebook is trying to create some new web 2.0 type of advertising model. It is trying to be different in the ad space and it is not working.
Although, I do think advertising can work for Facebook. They need to start small and focus advertising around features that people use on facebook. There are certain applications that people use more then others on facebook. Mainly photos, videos, events, news feeds AND SEARCH. People search on facebook, not so much to find information as they use google for. I think facebook can create a similar adwords model and build advertising around those core services.
Focus is the reason that search advertising is so wildly successful. If you’re looking to sell Budweiser, what better way is there to reach customers than to show your ads to the guy who goes to Google and types in “beer”? You know every single one of them wants to know something about beer.
Is this an automated response from adblock? Someone always says this in any topic about advertising. I'm starting to think it's adblock saying it, and not the user who decided to install adblock.
This is why I dislike Matt Maroon. I never met anyone who was so successful in coming across as completely arrogant using only text.
Advertising has traditionally been very unfocused. Anything that focuses it will work better than unfocused advertisment. Facebook has the data to focus advertisments. So neccesarily, Facebook works better in delivering advertisments to users. QED.
If or not they are actually executing this optimally is not relevant, the fact is that they have the capability.
Microsoft and a bunch of very smart business people are willing to bet billions on this. I think I'll stick with their opinions, rather than that of Matt Maroon, a very mediocre smalltime poker player.
He doesn't really explain how "advertising works," only how search advertising, specifically works.
So the conclusion of his argument should not be that Facebook won't make money off of ads but rather that Facebook won't be able to make money in the same way Google does.
The reasons why advertising doesn't work on Facebook are more subtle than the "there's no intent" line, which was trotted out by at least a dozen people last year. Google "scoble Facebook stroller" to see.
This is a good point, Facebook has 1 part of the puzzle, Google has the other. Facebook knows "who you are" better than Goog, Goog knows that you want "right now" better than FB.
I think that Google probably knows who you are pretty well. How often do you log into Gmail and then forget to log out and do searches on Google? Even without doing that, it's still tracking you via cookies.
You're fooling yourself if you don't think that Google has been tracking you and your searches for the last ten years.
In fact, I would guess that it knows a fair amount of things about you that you would never post on Facebook.
Google's model is nearly optimal in many cases, but I believe many advertisers still spend the majority of their money outside of Google. The reason is that there are two kinds of advertising, roughly: meeting demand, and creating demand. Google is good for the former ("I'm looking for X"), and TV has been great for the latter ("want to be important? You need this expensive car").