As someone who spends way more time on my computer than is probably healthy, I am familiar with online ads because they bombard me all the time. We know that online advertising has been a huge wealth generator for those companies such as Google and Facebook that sell and place ad space and indeed is the source of revenue for any ‘free’ content on the internet. (This blog is the exception. It is genuinely free!)
The promise that the media platforms offer to ad buyers is that they can target the product to the individual buyer who may be looking for that particular item, thus preventing wasteful blanket ads. The way the platforms do that is by vacuuming up all the information they can glean from us from our online activity to create a detailed profile of each one of us that they can then sell to retailers, a process known as ‘microtargeting’. This promise of ultra-efficiency is what has led the migration of ads from print media to online media. It has also led to fears that that we are now living in a Big Brother world because our lives have become transparent to these big companies like Google and Facebook who seek every way of prying into our lives.
But while all of us have received multiple ads for products related to any internet search query we made or something we bought, is the return for advertisers worth the cost? It is not easy to get an objective answer because measuring effectiveness was never easy even in the former analog world but it has become even more difficult in the digital world because there are so many extra layers involved. This is also a hot potato of a question because there are a vast number of people involved who need us all to believe in the efficacy of the system and there seems to be an uneasy fear that this is an ‘Emperor has no clothes’ situation where it might all be an illusion that could come crashing down if we look too closely.
We are in a situation where sellers of products and services feel they must advertise because their competition is doing so and they fear that they will lose out if they don’t. It is FOMO in the retail sphere. When people do something because others are doing the same thing, that has all the hallmarks of a bubble. So one can understand why the advertising platforms and the empire of ad makers and consultants may not even want to know the answer, fearing that it might prick the bubble.
And some are raising alarms that this might indeed be the case. Gilad Edelman has a highly informative article that reveals what is going on behind the scenes.
The real trouble with digital advertising, argues former Google employee Tim Hwang—and the more immediate danger to our way of life—is that it doesn’t work.
Hwang’s new book, Subprime Attention Crisis, lays out the case that the new ad business is built on a fiction. Microtargeting is far less accurate, and far less persuasive, than it’s made out to be, he says, and yet it remains the foundation of the modern internet: the source of wealth for some of the world’s biggest, most important companies, and the mechanism by which almost every “free” website or app makes money. If that shaky foundation ever were to crumble, there’s no telling how much of the wider economy would go down with it.
Hwang points to disturbing similarities with the subprime mortgages bubble of 2007.
Just as housing played an outsized role in pre-crash financial markets, so does advertising in the digital economy. Google earns more than 80 percent of its revenue from advertising; Facebook, around 99 percent. Advertising also makes up a fast-growing share of Amazon’s revenue.
If the financial market of the aughts was dangerously opaque, so, too, is modern internet advertising. In the early days of online ads, a brand would strike a deal with a website owner to host a paid banner. The onscreen space for that image, known as the ad inventory, would be sold by the publisher directly.
Today, the process has grown far more complicated, and humans are barely involved. “As they do in modern-day capital markets, machines dominate the modern-day ecosystem of advertising on the web,” Hwang writes. Now, whenever you load a website, scroll on social media, or hit Enter on a Google search, hundreds or thousands of companies compete in a cascade of auctions to show you their ad. The process, known as “programmatic” advertising, occurs in milliseconds, tens of billions of times each day. Only automated software can manage it.
Hwang says that what is driving the bubble is that most ad buyers do not realize the worthlessness of what they are buying.
There are piles of research papers in support of this idea, showing that companies’ returns on investment in digital marketing are generally anemic and often negative. One recent study found that ad tech middlemen take as much as a 50 percent cut of all online ad spending. Brands pay that premium for the promise of automated microtargeting, but a study by Nico Neumann, Catherine E. Tucker, and Timothy Whitfield found that the accuracy of that targeting is often extremely poor. In one experiment, they used six different advertising platforms in an effort to reach Australian men between the ages of 25 and 44. Their targeting performed slightly worse than random guessing. Such research indicates that, despite the extent of surveillance tech, a lot of the data that fuels ad targeting is garbage.
That explains a lot because so many of the ads I see on the pages I visit seem obviously irrelevant to me. The online bridge game site I visit has two side panels of ads and for the longest time, they have been advertising long, flowing dresses for both formal parties and summer casual wear. Of course, this may be because I am just the kind of person they do not care about and so I get random ads. They seek younger people with disposable income who can be persuaded to purchase things on a whim, not an old guy set in his ways who absolutely hates shopping.
There is also a lot of waste and fraud.
Then there’s the astonishing level of digital ad fraud, including “click farms” that serve no purpose other than for bots or paid humans to constantly refresh and click ads, and “domain spoofing,” in which a bottom-dweller site participates in ad auctions while disguised as a more prestigious one. Hwang cites a 2017 study finding that, between lousy ad placement and outright fraud, “as much as 56 percent of all display ad dollars were lost to fraudulent or unviewable inventory in 2016.”
So this raises the obvious question.
It’s fair to wonder why, if programmatic advertising is such a bum deal, so many brands continue to pour money into it. The reasons are manifold and overlapping. To begin, most of the people responsible for ad spending have no idea where their ads are actually running, let alone how they’re performing, and certainly have not brushed up on the latest research papers. That’s especially true for the small and medium-size businesses that make up the bulk of Google and Facebook advertising customers. I spoke recently with the owner of a successful online audio equipment store who had recently learned, thanks to a chance encounter with an expert, that 90 percent of his programmatic ad budget was being wasted on fraudulent clicks. Most other merchants simply never find out what happens after they send an ad out into the world.
How might this end?
So if Hwang is right that digital advertising is a bubble, then the pop would have to come from advertisers abandoning the platforms en masse, leading to a loss of investor confidence and a panicked stock sell-off. After months of watching Google and Facebook stock prices soar, even amid a pandemic-induced economic downturn and a high-profile Facebook advertiser boycott, it’s hard to imagine such a thing. But then, that’s probably what they said about tulips.
This is not something to be cheered. However much targeted advertising may have skewed the internet—prioritizing attention-grabbiness over quality, as Hwang suggests—that doesn’t mean we ought to let the system collapse on its own. We might hope instead for what Hwang calls a “controlled demolition” of the business model, in which it unravels gradually enough for us to manage the consequences.
Either way, the collapse of the bubble is not going to be pretty.
steve oberski says
I am familiar with online ads because they bombard me all the time.
A good ad blocker (I use AdBlock, available for Chromium and Firefox) will make that a thing of the past.
Also suggest uBlock Origin and DuckDuckGo.
And definitely a paid (not free) VPN service.
The entire field of advertising is a mystery to me, though. Before I’d get too worried about the entire internet going up in smoke, I’d want to see a study that compared internet ads to ads in other media — because advertisers won’t be fleeing the internet if they don’t have anyplace better to go. And it definitely seems as if brands that advertise do better than brands that don’t.
As you say, I wouldn’t mind being fed ads by an algorithm so much if the algorithm was better. I’m a gamer, so I generally find game ads interesting and often click on them. But if I buy a nozzle to replace the broken one on my washing machine, why do I get bombarded by ads for the exact same nozzle for the next month? Do they really think I want to buy a whole bunch of them?
I also think that microtargeting may not be the best strategy. You are selling only to a narrow group of people whom you think would like your product based on their past behavior. And, there might be something to that.
But, don’t you want your product to be blanketing everyone’s consciousness, even those who haven’t bought something similar yet? When people have to start buying something new, they inevitably drift to something they’ve head of. If you don’t buy tablet computers, then for some reason have to, for a present or work or something, don’t you look at models whose names you know? iPad, for example, not some other brand.
Or mattresses: I don’t have occasion to buy those often, but when I did, I turned first to Sears, because that name has been branded into my consciousness for decades as one that sells mattresses.
I just don’t think that frantic buying of data in such detail can give a good return on investment.
Pierce R. Butler says
Pls note that the Wired article exploring this issue dates back a year and a half; “Hwang’s new book” even longer.
Almost all the ads I see are for t-shirts and other clothing, mostly not for my gender, even before the one time in my life I purchased a t-shirt online last year (to support a local event, not because I saw an ad for it).
This year’s real-estate “bubble” seems driven more by megabucks speculators buying up housing in bulk for rentals than anything else; it may actually persist, in the absence of legislation to break open the market again.
steve oberski says
@4 Pierce R. Butler
What’s the difference between paid and free VPNs?
“Free” VPN services routinely sell your browsing and IP address data.
Also know to route you through proxies that redirect your URLs to bad actors.
I suggest researching the paid VPN services, you especially want one that advertises a no logging policy.
Mano Singham says
Actually, ads don’t bother me that much. My eyes just glaze over them. They also kind of amuse me in that when I see one that catches my eye that seems strange, I try to guess what internet activity I have recently indulged in that may have prompted that ad.
My Rick did direct marketing for magazines he worked for, back before Congress betrayed our founding father Benjamin Franklin by forcing the USPS to remove the subsidy for those.
Magazines were (my emphasis, were) the directed marketing for advertisement, a process massively overdeveloped by Tina Brown. If you got a specific kind of magazine and since postage was subsidized magazines could be very cheap, and there were a lot of different kinds of them once-upon-a-time. If you were a subscriber, and magazines Rick worked for who were more slimy would fake those, the subscribers were interested in that subject.
Mano, you are supposing that there is an alternative to what is happening for advertisers? Yet print media is largely dead. If I look at my local paper though my library’s link it does have advertising that’s built into the newspaper that my library doesn’t block, because it’s part of the ‘newspaper’.
For me, I have a Chromebook. For awhile, Google and I had an implicit deal for me using their OS that they would take my information as required by their EULA. In return they would (and still do) protect me from viruses etc. that are not theirs. They also used to prevent advertising that would slow down or corrupt the usual function of Chrome. Because of this I used to freely allow commercials/ads to save me the effort of having my Windows machine fail and corrupt my data permanently every six months, or zarking Apple “upgrading” every three years but not allowing any back compatibility pretty much wiping out my data without actually wiping it out because how the heck am I gonna find it now I am not a programmer?
It was a good deal.
Now, I have to use uBlock and NordVPN, as suggested by other users. AdBlock Plus got corrupted when it got bought out, it no longer prevents a Chrome machine from crashing/malfunctioning. Chrome still blocks viruses, so while the deal isn’t as good as it used to be, I persist in using Chrome.
More on topic: Chrome ads used to be fairly well directed at me, before I needed to use the uBlock NordVPN combination. The exceptions were small semi-personal sites run by one person just because they wanted to: they all got inundated by pornslime and semi-pornslime. More recently, I got an Android phone and before I turned off or disabled the microphone on all apps I possibly could, conversations I had with Rick or with people I had at the bus stop would result in relentless targeted advertising on Hulu.
So targeted advertising isn’t a complete fail. However, they have been declaring war on us users using up more and more of our computer’s capacity, and their selfish malice is fucking everyone and everything up ~~~to the degree that you or any of all y’all consider advertising a useful part of modern society, of course. c.f. Marcus about this also.
It’s a disaster of their own making. And yes, I remember in college those who went into marketing: they couldn’t handle anything else not even an english or psychology degree. It’s a bunch of personable persuasive people who are not pretty or well-voiced (otherwise they’d be actors, and some of the ones I knew in college were trying) but are so dang dumb.
steve oberski says
I guess tolerance for ads varies, I know that mine is zero.
An interesting experiment, try going ad free for a while and then disable your ad blocker and gauge your reaction.
I’ve had similar experiences with watching ad free TV (back in the day) and then (very) briefly being exposed to commercial TV, my response was immediate and visceral.
On the darker side ads are a well known vector for delivering trojans, viruses etc.
Mano @#6, third-party ads -- i.e. not from the original server of the site you visit -- can be a source of malware. I fully agree with steve oberski @#1. For Firefox, I additionally recommend the add-on Privacy Badger.
Or, just install the Brave browser:
Pierce R. Butler says
steve oberski @ # 5 -- Thanks!
I had thought the Mozilla VPN was a freebie, but now see they charge $5-10 per month depending on duration of subscription.
Years ago, Marcus Ranum avowed getting a VPN would -- guaranteed -- draw NSA etc attention. Do you think that still applies?
steve oberski says
Pierce R. Butler @10
A VPN uses an encrypted tunnel that establishes a secure connection directly from your computer (in my case my router, I use dd-wrt and run OpenVPN right in the router) and the VPN server.
In some jurisdictions (Canada in my case) your ISP is required by law to track usage and make this information available to various levels of law enforcement on demand.
This channel is no longer available if you use a properly configured VPN connecting to a reputable VPN service so agencies like the NSA would be forced to target your VPN provider, making their snooping much more difficult.
I’ve looked at statistics for VPN use and seen figures in the 50% range for the US, although a distressing large number use free VPN services, so the NSA would have their work cut out for them.
This is a very informative video about what VPNs can and can’t accomplish: https://www.youtube.com/watch?v=9_b8Z2kAFyY
In addition, some paid VPNs will, if so enabled, do ad-blocking at the server, so the junk never even gets sent to your browser (albeit using a local ad-blocker is also sensible, some browsers have fairly decent ones built-in, but it may need to be enabled).
In addition to the previously-mentioned Privacy Badger (highly recommended), another extension worth looking into is HTTPS Everywhere, which converts all
https://whenever possible. (Some browsers have such a functionality built-in, but it may need to enabled.)
Yes, using the Brave browser can also be sensible.
VPNs (and ad-blockers) can mess up some functionality. One amusing example is a VPN will foul geolocation (typically), which isn’t necessarily a bad thing. But, as some sites make assumptions — e.g., what language or currency to use — based on geolocation, there can be some extra hoops to go through.
And back on-topic: Do on-line ads work? Yes, in a negative sense: I deliberately avoid whatever product or service is being shoved down my throat using-up my bandwidth, etc. This includes spam (e-mailed adverts). So if you don’t want me to even consider being a customer, then either manage to present a totally-unwanted ad, or insist I disable “the” ad-blocker. (There are various exceptions, e.g., I’m already using the service (so why is being “microtargeted”-at-me?). Or, for a few sites, I concur to sign-up for a newsletter, but typically use a secondary e-mail account to help keep spam away from the main account.)
It’s the sodding ads on YouTube — the ones that interrupt the video — which are, at the moment, infuriating me the most.
For me, the thing that makes “targeted ads” most laughable is the programmers’ obsession with key words, as if that has any bearing on the person. We’ve seen that as atheists, ads for religion dumped on us because we discuss it. People in anti-car groups report they often see ads for new cars, but not a single ad for bicycle companies or train travel, things that would actually be of interest to the group members. Childfree groups say the same thing about seeing products for kids, daycare, etc., advertised to people who will never have kids.
One big change I have noticed in targeted ads is language. After I moved to Asia, ads were targeted geographically and linguistically, to match the local population. They were never readable to me, and I ignored them. But over the last year or so, the majority of ads that I (not friends) see are in English, though that doesn’t mean their targeting is all that good. A lot of them are for “US green cards!” or “study in Canada!” Not exactly something that would appeal to me.
Deepak Shetty says
A positive of online advertising (for the marketing department) is that you can *usually* figure out if the user converted i.e. they bought the thing being advertised -- And this can be done (Due to the various tracking tools like cookies) across sessions , across websites. Ubiquitous tools like Google also probably know whether you considered alternatives and what prices you saw at the competitors etc.
There is also the cost of not advertising. Search “Apple iPhone” for e.g. -- Google knows enough of that search term because it can show me information about “Apple” the company on the right hand side and the Apple website as the first result. But the first half of the screen is Ads for Shop for iphones by various providers which i can shrug and live with. But the funny thing is there will be an AD by Apple itself! And most companies have to do it because Google the advertising company has ensured that if you dont do this Apple’s competitor can then get an Ad before the highest ranking result with whatever text they want ! Almost all companies have to buy their own keywords even if they rank top. If I search for “$Product name” “My Company name” -- I invariably see my companies competitors ads even though its quite clear to google that my intent is my companies product .
Some of the childish algorithms make me laugh. I browse my companies site because that is my day job. Whenever I goto websites related to my hobbies I get ads of my company , no matter how many times I dont click anything. The subject matter of my hobby being completely unrelated to the subject matter of my company never seems to stop these algorithms.
Microtargetting may be getting superseded by the brainless ML-AI type of ad schemes
I haven’t had any success with display ads. Even if your target might be interested someday, it is unlikely they are today. So, it’s more about brand awareness which can be hard to quantify for a small company. Search advertising -- the ads that appear at the top of the page after doing a Google Search -- can provide a positive ROI depending on the product. Search “my roof is leaking” and you will probably have several local pay per click ads appear at the top (assuming you aren’t hiding your IP). If I owned a roofing company, I would be extremely interested in my ad appearing to that searcher.
The most productive advertising is still referrals/word of mouth.
John Morales says
marner, you’re damning with faint praise.
“Can” being the operative word.
Of course, the overwhelming majority of online advertising is other than the ads that appear at the top of the page after doing a Google Search.
Steve Morrison says
It’s also possible to block ads using a HOSTS file; suitable files can be downloaded from various sites.
Marcus Ranum says
Years ago, Marcus Ranum avowed getting a VPN would — guaranteed — draw NSA etc attention. Do you think that still applies?
I know it does. But nowadays they collect it all and the analysis is done by AIs. Unless you’re someone politically significant or you are doing crimes or kid porn you’re just another little drop of data in the tsunami and you’ll most likely attract no attention. You should assume your traffic is safe from marketers, but I would not recommend trusting a VPN to do crimes.
I personally don’t really do ads. I think there are only two kinds that actually get through to me anymore. The first type are those on youtube that are directly sponsoring a video. And often I’ll skip those so the ad buyer is mostly just getting name recognition and little else from it. That’s not nothing but it does mean they’re sometimes buying a 20-30 second ad slot detailing their offering but for me they’re only getting 5 seconds of “this video is sponsored by X.” Sometimes the video creator will do an actual skit themselves to work in the sponsor’s product. Those can be entertaining and I tend to see more of whatever they’re doing. The second type of ad I see are emails from companies I actually buy things from regularly. These are things I shutdown if they aren’t something I care to see.
Everything else I block. From an adblocker that’s already been mentioned to privacy badger blocking some trackers to a list of around 175 domains I block at the router. That list has been growing for around 20 years now and is a list of domains that hosted ads placed on pages with content I wanted to see. It started when I had much slower, older computers in college and was dealing with this stupid flash ad that would eat all my resources and make noise. It was one of the first that tried making noise to draw attention. It had a mosquito in a banner ad and you were supposed to click it to try to “squash” the bug or something, I think. This was years before youtube and browsers were not terribly sophisticated about media, so I had to dig through tabs to find the damn thing and close it while the ad was slowing my system down significantly. I think ad blockers were primitive and also slowed down my older system, so that’s why I began my list.
And that’s basically it for ads. They barely exist in my world because they rarely respect my time and they won’t respect my interests even if I let all the trackers through and wilfully cooperate in exposing my interests to all the algorithms. And I don’t see any reason to expose myself like that, I’d rather donate to the Electronic Frontier Foundation (privacy activists who also make privacy badger and many privacy guides among other things) instead.
Oh, quick caveat: I worked for a middleman that sold better targeting to companies. While I was there they were just adding in programmatic ad buying. This was a few years ago. All I did for that company was make sure their servers didn’t fall down and help with tech support issues surrounding the servers. But it did let me get a glimpse behind the curtain a bit.The truth is I didn’t learn anything that changed my mind or meaningfully altered anything I heard from the EFF. It just provided more perspective and made me even more certain I was acting in my own best interests.
I recently used a friend’s laptop and was appalled by how many, and how intrusive, the ads were. I had to dig down through pop-ups and hunt for how to close full-screen ads. It was damn-near impossible and I was dearly tempted to toss the laptop into the trash. I settled on gritting my teeth and doing the bare minimum.
bluerizlagirl . says
There’s a really insidious aspect to Internet advertising, in that adverts can be very precisely targeted so that they are only seen by the desired audience and not by anyone else.
With traditional advertisements, everybody sees exactly the same advertisement. If you publish a racist (or sexist, or homophobic, or ableist, or transphobic …..) advertisement in a newspaper, on a billboard or on TV, everybody will see it and some of them will complain. This is a simple social filtering mechanism, and it lets bigots know their bigotry is not acceptable.
But on the Internet, you can choose exactly who gets to see your advertisement and who doesn’t.
If you publish a racist advertisement on the Internet, and make sure only racists get to see it, then the racists who see it will naturally assume that the lack of complaints from normal people is because normal people don’t think it’s racist. It doesn’t occur to them that other people might not be seeing the same advertisement, because they are so used to the way traditional advertisements work.
By carefully targeting advertisements away from anybody who might complain about the content of them, you can manipulate people into believing an idea has more support — or at least, less opposition — than is the case.
Show them images of people like them or maybe a little bit older (to suggest things used to be better) in a superficially familiar environment with some slightly-exaggerated features, openly using slurs and making outrageous allegations that go unremarked-on, and you might be able to persuade them that minorities pose an existential threat to them. And in times of existential threat, altruism is abandoned as we switch survival mode from “gregarious predator” to “solitary predator”: one fat caveman has a better chance of passing on his genes than two thin cavemen. Conversely, people are much keener to share what they know is plentiful; in times of plenty, the pack instinct takes over, and it is easier to show concern for the population as a whole. Doom and gloom are good for business.
I think that’s the real danger of targeted advertising, and I’m not even sure whether it’s possible to do anything about it. There isn’t a technological way of preventing it, and criminalising it would not work because it’s so hard to prove it is even going on. And it doesn’t matter how a crime is punished, if it goes largely undetected.
Pierce R. Butler says
Marcus Ranum @ # 19: … I would not recommend trusting a VPN to do crimes.
Good tip, thanks. I quite prefer the more traditional methods of blunt instruments and dark alleys.