The Facebook Fallacy

It's not that internet advertising isn't effective. It's that all the content it's wrapped around is too engaging.

The reason ads in real-life are effective is that they traditionally haven't had to compete for your attention. Examples:

(1) Standing on a subway platform you are looking straight at a piece of advertising for minutes until your train arrives. (2) Driving down the road, you're staring at a billboard for 1/4 of a mile (15 seconds at 50mph). (3) On TV and on the radio, inertia to change the channel means you often listen to ads.

The problem for those media is that we are now reaching the point where content is portable enough that we now never have to be subject to ads in those public spaces because we can direct our attention to our mobile devices. Over the next for years you're going to see a massive drop in the effectiveness and reach of ads in real life because people will no longer be directing their surroundings.

The only one of the above examples that will survive a while are billboards and even they will only last until we have self-driving cars. Once we have self-driving cars, why even bother looking at the road instead of our iOS and Android devices?

Good of you to submit this with the canonical URL. All four of the previous submissions (none with any comments, per HN Search) were of noncanonical URLs with appended referrer strings or format-preference strings.

After edit: So the article's essential claim, which I guess we haven't discussed here on HN before (unless I missed a link) in relation to this exact article is that the Web is a fundamentally unfavorable environment, long-term, for advertising-supported sites. That's an interesting idea. I usually run ad-blocking software, so maybe I believe this already. I've never tried to monetize my personal website with advertisements--they just seem too schlocky.

My hope is that advertising, as an industry, dies (or morphs) sooner rather than later.

At best, it is visual pollution, and at worst employs subtle forms of mind control to encourage people to do what the advertiser wants. No thank you!

However, it is useful: it helps disemenate information. Didn't know about the all new BMW i3 electric car? Thanks to some ad, now you do. Most of the time you don't care, but when you do care, it can be useful.

The Internet is much better at dissemanting relevant information than advertising, however. Why should I watch an ad for a movie if rottentomatoes.com has all the info I need? Why look at car ads when websites will happily tell me the pros and cons of various cars? My hope is that eventually the best products and services for me will be peer identified an given to me when I need them. Ads will no longer be necessary. The peer review process must be paid for somehow; manufacturers may end up footing the bill, or perhaps ads from unpopular products will; either way I cannot wait for this garbage/propaganda called advertising to vanish from the earth.

The key problems for online advertising are:

1. Advertisers can see if it worked. In traditional advertising you accept that a lot of your money will be "wasted" on people who will never buy your product or service. In online advertising you can track a customer from first viewing an ad to purchase. This gives advertisers a lot of leverage in price negotiations -- "your site only has a 0.05% conversion rate, why should I pay that much?"

2. Inventory is growing faster than demand for inventory. In advertising terms "Inventory" means "I have a place to show your ad". Well every new web page can contain inventory. Given how fast the supply of webpages is growing, compared to the total pool of customers and businesses ... well let's just say that the supply-demand curve on this does not favour the sellers of inventory.

I am of course fascinated by this problem and my startup (currently in stupid unfunded single-founder mode) aims to resolve this exact economic problem with a different model.

Here's the core problem as I see it: the old (advertising) math no longer works, and that's getting people into trouble. For traditional advertising in print, radio, or television the math is simple and it's been refined over decades. Generally speaking you take the number of eyeballs or ears, you multiply by a certain factor which takes into account things like demographics and whatnot and then you're done. In large part it doesn't matter whether someone is watching murder she wrote or magnum PI or reading the paper, for each format there's a degree of effectiveness of advertising and a degree of lucrativeness for the target demographic. There are of course complications, but in broad strokes this is the way things worked.

Those assumptions completely fall apart on the web. Because the web is not just passive. It's interactive. And that changes the equation tremendously because it means that the effectiveness of ads can vary by an enormous degree depending on activity. When people are searching, for example, relevant ads (not just random eyeball hijacking brand promotions) are actually useful, and thus highly effective. When people are trying to communicate with their friends there probably isn't a good advertising context at all most of the time and almost all ads will be seen as useless interruptions and thus vastly less effective (they won't be relevant and can't as easily be made relevant). And there is a huge variety of other activities on the web that don't fit into convenient molds.

Now, does that necessarily mean that advertising revenue is a thing of the past? Obviously not, as in some cases there are circumstances where the potential for effective advertising is even greater than with traditional media. But it does mean that the math is far more complex. And it might mean that just because you have a direct line into a huge amount of eyeballs you may not be able to monetize that traffic to the same degree (by orders of magnitude) as an equivalent amount of traffic of people engaged in a different activity.

Additionally, we've seen the writing on the wall in terms of "eyeball hijacking" highly interruptive advertising in the modern age, it's just out of date and out of touch. Monetizing web traffic effectively requires a much smarter approach. In the ideal case advertising is so well targeted and so effective that it is welcomed and often not even seen as an ad.

Advertising is a $600B a year business, and the Internet still has yet to catch up with other media in terms of share of total budget versus time spent on the medium.

Google just reported revenues up 35%. It's all advertising, driven by high quality algorithms and lots of data. Internet advertising is more effective than other media and we're still just scratching the surface.

I'd like to comment on all the specific ad startups and their revenue and growth rates but I'm not able - you'll just have to wait and see it.

Happy to make a $1 bet on the subject with any naysayer here or with the author of this shoddy article.

I don't like ads (typically). But as ads get more and more targeted (ie., when i haven't gone to the dentist in 2 years and the ad says, "great dentist with promo offer just 2 miles from you", that's kind of enticing). I think it will migrate to where ads need to be helpful (ie., google ads). Search was easier because Google could target on the search term. But it's harder when the user isn't proactively searching for something.

But I think it can be done. The ads just need to be super-targetted and helpful.

I don't understand why people are so focused on Facebook's ad revenue and it's supposed unsustainability, since Google has been around longer, are way bigger, and you know what? The overwhelming majority of their income is also from online ads.

And you know what? Before that we had TV, Radio, and newspaper ads. Those industries didn't implode (or start to) until better ad delivery systems (the web) showed up?

So you know what's going to kill Facebook (and google)? Not some realization ads "aren't working". A newer and even better ad delivery system. (also unlike old media, there doesn't yet seem to be a reason to say FB or google won't be behind this new system either).

This is tired old bollocks.

Here's the thing: I hate ads. All forms of ads.

The reason I hate ads is that 99% of the time are interrupting me and telling me about things I do not want to know about. If I am chatting with a friend on Facebook, there's no way any kind of ad is going to do anything but get in my way. If, however, I am actually looking to solve a problem, I go to a search engine and start bringing up web pages. In this scenario I am actively seeking something -- perhaps something an ad might be able to help with.

So content ads on informative sites and as part of search seems to me like a lesser evil of all ads. I have a few sites and ideally I'd like to remove all Adsense from my site and replace it with links to stuff people want. The problem is that there's such a long-tail phenomenon going on when people go looking for stuff. For instance, if I have a informative site on dinosaurs I know people coming there via a search engine want to know more about dinosaurs. But would they like a book? A museum trip? Tickets to a dinosaur movie? A college education in paleontology? The answer is yes: some of them would like all of that stuff. And even more. The only way that I know how to assist these people is somehow rotating ads that solve a large variety of problems. Ideally you'd use some kind of machine learning to gather what you can and continue to optimize. Google does this for me, so while I hate ads, providing people with a _little_ bit of randomized assistance from a large catalog of options in a location where they are already looking for stuff seems like a pretty good deal for both the reader and me.

In social media you are no longer trying to respond to a searching reader. Instead you have readers that are doing other things that you are trying to interrupt. Perhaps using their friends as interruptions might work. Perhaps if you only interrupt by providing material related to what they are already doing. But no matter how you hack it, as a social media advertiser your game is to take a reader that is doing one thing and get them doing something else. We've gone from ad-supported radio and television, where listening to the ads were a form of saying "thank you" to the advertiser for providing the material, to a place where advertisers are using technology to try to pry into and control every part of our lives. And yes, it's because of everybody wanting things for free. This is not a happy symbiotic relationship anymore.

ADD: In fact, it's mind-boggling if you think about it. Facebook is trying to take a billion people and make them do things they would not normally do. Yes, people waste their time anyway, but not everybody, not all the time. Wonder how much lost productivity are they creating? How many man-years? How many people making purchases for things they never wanted? How many people playing games with imaginary farms and such for hundreds or thousands of hours? People said the television had an on-off switch, and so does the net. But people are social animals, and the net is the way we communicate. If all of your friends and family are on Facebook, you're going to want to know what they are doing. Facebook uses your friendships against you in order to purposely divert you from things you'd normally do so that you'll spend all of your time on their site playing, gossiping, and purchasing. Ouch. Doesn't sound like an imminent failure. Sounds like a new form of crack. If that much social damage were done by a drug, it would have easily been outlawed by now.

I believe I already read this article a while back on HN: http://news.ycombinator.com/item?id=4007047

Anyway, it raises a very important point, not just for the advertising web but for IT in general.

it seems part of the catalytic nature of IT, that it does both increase efficency and reduce effectiveness of ads. because the increased exposure reduces the effectiveness in our minds, the omnipresence of ads online will cause their prices to fall, causing the need for more ads. it's like a death spiral. however I'm wondering then how could print and tv ads survive for so long? maybe one solution might be to actually decrease the number of ads on the web - to maybe one prominent, shortly displayed ad per page (could there be a study measuring the average screen/page share for ads in print and online media?).

or if that fails: IT costs have also been decreasing every year, hence even if facebooks revenues might stall, ever lower server costs might allow them to keep a profit. nowadays a one man show can in many instances run a small business on the web with thousands of clients.

I think facebook should try a groupon style model. They have locations and intentions of users. They should just have a deal feed, or some type of email listing with possible low prices they could get if they join in a deal. To filter it better, maybe ask each user for a list of things they want cheap.

It makes me sad that so many people see the internet as just a tool for advertising.

At least there's ad-block, I guess.

1. Install Trueblock plus + cookie monster and disable third-party cookies. 2. You just broke google/facebooks business model. Haha.

goog/fb are relying on peoples stupidity and not going to tools>adons and installing ad-blockers.

Why can't Facebook start charging for its API, and for business pages? Banner ads were big in the 90s before people realized they were that effective. I don't see why Facebook would need to rely on them forever.

Facebook should look at doing financial services. Price credit based on who your friends are. If you don't keep up with repayments you get a permanent blot in your profile. It is a sector in need of disruption, and Facebook have the worlds best dataset for assessing risk. You could start out with small transfers between individuals. Then move on to group based credit card rates.