A variety of media stocks may fall even further

…Google announced that 56.1% of ads served on the internet are never even “in view”—defined as being on screen for one second or more. That’s a huge number of “impressions” that cost money for advertisers, but are as pointless as a television playing to an empty room.

This is not a big revelation. The web metrics company ComScore reported last year that 46% of online ads are never seen. Spider.io, an ad fraud company acquired by Google in February, has pointed out that a large portion of ads are “viewed” only by robots, revealing that one botnet of 120,000 virus-infected computers viewed ads billions of times, running up the tab for advertisers without offering them the human eyeballs they sought.

There is more here, by Zach Wener-Fleiner, and for the pointer I thank a loyal MR reader.


Never-the-less the ad-based web economy is the engine of the future and it will prove to be the source of more wealth generation than anything else in history.

Poe's Law?

No, spoof of another regular poster (screen name: ummm)

By Jove! Keep on the scent Sherlock!

I like this version of you so much better, so I'm glad I drew you out of whatever cove you were hiding the HMS Mercatus in.

We need more immigrants to see online ads; that's a job Americans just won't do. Right?

The price of web ads reflects their less-than-stellar performance at generating incremental sales lift. Per-impression, the cost of a web ad is about one-tenth the price of other channels, which is about the ratio of their effectiveness on the margin. If the technology improves to reduce this type of fraud, then the prices of this ad inventory will increase. Unfortunately, the direction of digital ad sales is towards more automation (called programmatic buying and selling), which yields even more opportunities for fraud to enter the market putting further downward pressure on prices.

@bastiat - CAPTCHA filters before you can view the online ad? lol

BTW all the greats of yesteryear are only great today due to anachronistic thinking: Socrates was considered an idle stonemason and state subversive in his lifetime (his student Plato elevated him to fame); Bastiat had little or no influence, as the state was not that developed in his time; Keynes, borrowing ideas from other proto-Keynesians, was not followed much in his lifetime and his Bretton Woods agreement is just a compromise to maintain a gold standard (fixed exchange rates backed by gold), and so on. Only afterwards do future generations elevate these people into profits.

As for me? You mark my words, if these comments are preserved future generations will marvel at my words on increasing innovation through patents (AlexT has an interesting book on IP, that starts off and ends good but strays into an anti-patent discourse: "Launching the Innovation Renaissance".

The best minds of my generation are thinking about how to make people click ads. That sucks. Now they also think about how to make people view ads, how to prevent other people from avoiding viewing ads, and how to prevent still other people from pretending that some people viewed ads when they really didn't. That sucks squared and cubed.

But I'm smart, so instead of thinking about how to make people click ads (and all that other stuff) I sell people software that helps them think about how to make people click ads. That's much better, I can donate to open-source projects, angel-invest, support social justice and wear panda hoodies.

Made from roadkill pandas, of course.

I'm generally wary of people who say that smart people are such an important resource that we should smash their livelihoods to get them to do something more useful. If you want the "best minds" to do something else, then pay them better to do something else. But as far as I can tell plenty of smart people still work at the LHC or NIH. In fact the median age of physics postdocs is 38, and its not different for most other scientific fields. So it's not exactly like would-be cancer curers are out there offering great pay and jobs to young, bright, ambitious people.

No, they're not offering, which is exactly the point. But I'd like to see you apply this argument to such moderately smart and ambitious people from disadvantaged backgrounds as those who become drug gang leaders (what was the name of that Indian PhD student who wrote a thesis on this?) They earn plenty. Should we smash their livelihoods to get them to do something more useful?

I agree we should not force people to become drug gang leaders. We should legalize drug selling so that people who sell drugs to willing buyers operate within the legal system and don't need to have their own private enforcement systems. That is legislation is creating the problem by the unintended consequences route. Similar to what would happen if you prevented "smart people" from developing advertiser technologies.

On a related note, probably about 90% of human activity could be described as socially unnecessary, in that it is not advancing the human condition. Why pick on advertising particularly? Myself I am very grateful for the advertising supported technologies such as the vast amount of free content and services on the web, not to mention the last 200 years or so of newspapers.

Referring to sociologist Sudhir Venkatesh?

I dislike this "the best minds are thinking about how to make people click on ads" quote.

I have never heard someone suggest the work of journalism is about getting people to respond to ads, nor the work of TV writers and news producers and such, though all of them are funded by advertising. Intuitively, people understand that newspapers and magazines and television networks devote their talent to more than just demonstrating the value of their ads. It is a gross ignorance to not appreciate that internet services do likewise.

But MPS, it really is very simple: the best minds don't work in journalism or in TV at all. If you have an extra good non-mathematical mind, I believe you end up as a super-expensive lawyer usually.

"The best minds of my generation are thinking about how to make people click ads."

Perhaps, but I wished they'd get better at it. I hardly ever click on an ad, except by mistake.

The other post about Apple vs Facebook should be pointing to this comment.

I'd rather have smart people working to make something as awesome as possible as cheap as possible because they will get more of my money, instead of spending their whole day putting me through a Skinner box to extract all my money.

Are TV or print ads really any more effective? Or is it just that the nature of the Internet allows the response to be better quantified. Apparently DVR has no impact on the effectiveness of television ads. The most reasonable conclusion is that TV ads have zero effect to begin with (0*X = 0).


+1. I worked for a few years in marketing and was amazed at how much money is spent on advertising without evidence as to efficacy. Of course, much of my job was to track and provide this evidence that marketing worked. (There was little incentive or desire to question whether it worked at all, because, well that's not good for the marketing dept.). If it does seem to work, it often seems to be in a small way w/ a high marginal acquisition rates. Much of the money is spent in a hurried and uncritical way: there's simply a budget, a growth imperative, and so it gets spent. But at least if we can't see results fall to the top or bottom lines, we can say we raised brand awareness and consideration - taking the pressure off of tangible results.

I managed a lot of BehaviorScan real-world lab test market studies three decades ago. Increased advertising worked mostly if you had some important news to convey about an improvement in your product. Otherwise, just spending more on consumer packaged goods brands didn't move the needle. Of course, these were famous existing brands that had immense amounts of recognition already from past advertising.

Seeing Michael Jordan eat a Big Mac on TV always made me go out and get one immediately, knowing that he preferred them to meals at his own restaurant.

I bought a newspaper for the first time in a year and was amazed and delighted by the print ads. They were so attractive-looking compared to Web ads and they weren't just targeted at selling me things I'd already thought about buy. They opened my eyes to interesting things I'd never imagined buying.

And newspapers ads are terrible compared to slick magazine ads.

What's the problem here?

If robots are going to become producers, surely they can be consumers too.

The value of an item of which there is an infinite inventory tends not to be every high.

Note that the article, based on Google's data release, mentions on-line ad IPOs may be inhibited (if they're competitors, good for Google) and buyers may be interested in Google's new product which only measures watched ads. This article seems like an ad for Google, disguised as information -- the best kind of ad, if the reader doesn't notice.

I meant buyers of ads, not consumers.

Clearly privately wasteful, but socially useful! Oh, thank god for the ads I've never seen... . Perhaps they paid for one or another thing I wished to watch.

The impression by eyeball model of how and why people buy is so 20th century.

It's really funny that this is what's essentially funding most free content.

This quartz article is pretty useless. It gives no indication of how many advertisers actually pay based on impression. My sense is not many. In affiliate marketing, for instance, commissions are based on sales. And in many other forms of digital marketing there's a more quantifiable action involved like click through.

Btw, when have the best minds ever been concentrated in a single field? I say enough with that nonsense.

I'm an engineer at an advertising analytics company. Publishers and advertisers that work with us get metrics like (distributions of) mouse hovering over ad time (we generate cute little heat maps for every creative), in-view time, % watched for animated and video ads, etc. Click-through rate is downright barbaric by comparison and I expect most companies who give even a little bit of a damn to work with us or a competitor within a few years.

I still find myself humming ads from the 70s and 80s. A good ad can still bring in returns 30-40 years later. One problem with internet ads is that they aren't very creative, even compared to print ads. Yet even crappy internet ads can work if they're well targeted. There's still a lot of room for improvement.

Google ad buyers pay per click-through, not per display.


I liked Mr. Cowen's blog entry.

does a blogpost exist if no human ever reads it?

As a daily web browser, I'd say I spend roughly eight hours on the Internet. While browsing online, I see several ads. With that said, I only listen to the ads that are unavoidable, such as ads on Pandora or Youtube. I close them out or skip them the second I have the chance to. Those unavoidable ads are the only ones that have a lasting impression on the surfers' minds though. The pop-up ads are essentially a waste of companies' time and money because it's wasted on viewers who wait for the 'X' to close them out. While I understand the need for advertisement, there are more affective ways to get messages across.

Comments for this post are closed