Carcinogens
If Edward Snowden was injected with a megadose of Super Soldier Serum, he’d look something like Frances Haugen. Perhaps Haugen’s disclosures — that among so many other evils, Zuckerberg knew Facebook’s products “harm children” — means that Facebook has crossed the wrong cowboys, specifically … cowgirls who are moms. MADD (Mothers Against Drunk Driving) finally galvanized the nation against the scourge of drunk driving in the 1980s — will MAMS (Mothers Against Mark and Sheryl) bring down the Zuck and his merry band of mendacious fucks?
But that is not what this post is about.
Facebook — all social media, really — is the nicotine, the dopa drip of outrage and baby pictures that keeps us coming back for more. But the carcinogen, the thing that should have warning labels slapped all over it and congressional hearings devoted to it, is … an algorithm-driven advertising model.
Ad-supported media has a long history, and it’s not all bad. Alcoa paid for Edward R. Murrow’s airtime, Woodward and Bernstein’s Washington Post relied on advertisers, and on Sunday night, Frances Haugen waited patiently for a Jack in the Box ad to run before she stepped out of the 60 Minutes phone booth in her superhero cape.
Even in traditional media, advertising has always been a problem — what stories did Murrow avoid while Alcoa was paying the bills? But on digital media, advertising has more potential and more power, and it corrupts the media businesses that rely on it. Digital advertising has exploded; even after a Covid-19 dip, it accounts for nearly half of all U.S. advertising spend.
Big Little Lies
This torrent of money is what fuels Facebook, YouTube, and the rest of the teenage-dystopia-industrial-complex. The transformation of media into social media into a surveillance-based attention economy is a direct result of the digital ad model. But there’s a second externality, and while it’s historically received less attention than the ill effects of algorithmic enragement bias, it’s a problem that’s grown in the shadows into a multibillion-dollar beast. Fraud.
The digital advertising industry is a Rube Goldberg machine of platforms, agencies, exchanges, and other middlemen. I’d explain it to you here, but a) I don’t understand it, and b) you don’t want me to. The number that matters is 89%. That’s the percentage of dollars spent on “programmatic” advertising. Ads bought by algorithm.
In a programmatic ad buy, the client — Nike or Nissan or Novartis, acting through an agency, the first of many middlemen — provides the ad itself and sets up criteria for who it wants to see it (e.g. 36- to 42-year-old Hispanic males with Crohn’s disease in the final year of their auto lease). Then a series of automated processes place many thousands of copies of the ad on many different websites, anywhere the algorithms believe the ad will be seen by people meeting the target profile.
That’s lots of palms to be greased. Lots of opportunities for people to cheat, and enough complexity that this cheating is difficult to detect. Especially if the cheating only makes the system more money.
The basic cheat is the fake view. An ad is reported as being served to humans, when it was actually only “seen” by a bot, or by a person in a “click farm” tapping at dozens of screens, or by nothing at all. Networks of fake websites fool the algorithms into believing they are real publications. Measurements of the impact are all over the map, but we know fraud is pervasive. By one estimate, 88% of digital ad clicks are fake.
Publishers and the middlemen who place ads with them tout all sorts of supposed fraud-detection technology, but industry experts say it’s largely worthless. Of course it is. These players benefit from inflated ad views — why would they suppress them? In 2008, Newsweek Media Group infected its own fraud-detection system with malware so it could charge advertisers for bot-generated traffic on some of its websites. Recently collapsed Ozy Media was a heavy buyer of fake traffic, and we haven’t seen the last Ozy-type scandal.
Even ads that do make it to real humans are not all that likely to be seen by the people the advertiser is looking for. This was the core promise of digital advertising — saving modern-day John Wannamakers the half of their ad budget that was wasted on uninterested consumers. But there’s increasing evidence that this promise was the biggest fraud of all.
A study by MIT professor Catherine Tucker found that even targeting something as basic as gender was unsuccessful more than half the time (i.e., it was worse than random). A Nielsen analysis of a household-income-adjusted ad campaign found that only 25% of its ads were reaching the right households. As much as 65% of location-targeted ad spend is wasted. Plaintiffs in a class-action suit against Facebook have alleged its targeting algorithm’s “accuracy” was between 9% and 41%, and quoted internal Facebook emails describing the company’s targeting as “crap” and “abysmal.”
And the technology that enables even this lousy tracking, the digital cookie, is on the way out. Cookies are short pieces of code websites leave behind on your computer so they can follow you across the Internet. But one adtech firm found that 64% of its tracking cookies are either blocked or deleted by web browsers. Apple recently updated iOS to require would-be ad trackers to obtain a user’s permission before dropping a cookie. Google’s Chrome (which commands 60% of the browser market) will block third-party cookies altogether by 2023. Although that cloud has a dark lining: Google is replacing cookies with its own proprietary system that will centralize ad tracking under its exclusive control. What could possibly go wrong?
Google and Facebook are the dominant players in this business. They’re huge publishers (capturing over half the digital ad dollars), and also the leaders in many of the upstream categories in the programmatic ad infrastructure. Google, for example, owns the largest digital ad marketplace, DoubleClick Ad Exchange.
Fraud is rampant in other areas of the digital ad business. Influencers can buy fake followers by the truckload — roughly 20% of them are fake. Approximately 40% of Donald Trump’s followers are likely bots. Social media platforms are rife with cats and bots: Facebook admits to shutting down billions of fake accounts on its platform every year. Even app store installs are fake. Bots/click-farmers download 1 in 5 iOS apps. On the Android platform it’s 1 in 4.
This problem pales in comparison to Facebook’s rage and confirmation bias debacle, but it’s still a serious economic issue. Criminality is a cost ultimately borne by consumers, and crime begets crime. Digital ad fraud could be a $150 billion business by 2025, which would make it the largest criminal enterprise after the drug trade — and it fuels the same digital criminal underground responsible for industrial espionage, ransomware, and identity theft.
We need externally imposed and enforced industry standards on transparency in advertising. Expecting these conflicted middlemen to self-regulate is (generously) naïve.
And we should consider taxing algorithms that serve ads and content. We tax cigarettes and alcohol to suppress their use and fund policies to address some of their externalities. Programmatic ad buying, similar to other media buys, can be good/bad, and that’s a component of business. But this is addiction, and it’s hurting all of us. It’s time for an intervention.
Don Draper, RIP
That’s what should happen. What will happen? The edifice of digital advertising is unstable and likely to collapse. The promise of measurable ad spend has been crack for chief marketing officers. That algorithm-driven media was destroying our commonwealth by accelerating the spread of misinformation and division wasn’t enough to give them pause. But now that they’re realizing that promise was a lie, some are putting down the pipe. Several large advertisers have made deep cuts in their digital ad budget — including Procter & Gamble (cut $200M), JPMorgan Chase (slashed ad reach by 99%), Uber (cut $200M), and eBay (cut $100M) — and seen little or no measurable impact on their business. Other large companies are building programmatic ad capabilities in house, figuring they can trust the tech if they built it.
These firms already have huge brands and global distribution, partnerships, and other means to sustain awareness. It’s possible to build a strong brand without advertising. Tesla is the most recent example, but not every company makes a revolutionary product in a highly visible consumer category.
Digital advertising promised small and mid-size businesses a way to take a small ad budget and increase efficiency. And where do these businesses feel they have to spend their money? The Facebook and Google duopoly. Which brings us back to where we started: mendacious …
Life is so rich,
P.S. Occasionally companies go viral because of what appears like luck. But in the Netflix documentary that gets made years later (hopefully a positive one), it’s clear that very smart people engineered that viral moment. Be like those people. Section4’s Science of Virality, taught by Wharton Prof Jonah Berger, closes in less than a week. Join us.
Love the messaged of this article, and was about to share with my digital marketing team, but first decided to fact check the numbers:
1) 65% of location data is wrong – actually said 29% was wrong and 36% they couldn’t verify. To me, the number 29% therefore are PROVEN incorrect, not 65%. Missleading
2) 88% of clicks are fake – when you read the source, the company goes on to say 98% of Google Ads clicks are bots. While I can’t verify this is false, we run Google ads and get conversions on well over 5% of ads. Their methodology seemed extremely poor, and I find it hard to believe the 3rd largest company in the world is selling 98% fake clicks. Can’t prove it, but the number seems absurdly high.
3) The demographic data makes the most sense, and I’ve often though that there was something wrong there. This one I believe.
I guess my point is to read more than the headlines, and I expected a bit more diligence from Prof G, but I guess he wanted to make a point
If you are correct, are the revised numbers referred to in your comment, since they’re lower than those reported by Scott, acceptable? No one in the C-Suite can come to that conclusion if they take the interests of their shareholders seriously.
great that you fact-checked the numbers yourself
The “numbers” Scott reports on between waste, fraud, bots, poor targeting, etc. are not new. They’ve been reported before and have been supported by numerous studies. They’re undeniable. Yet the problems persist while billions are being invested in programmatic advertising (and rising). My question is a pretty simple one. When will CEO’s, CFO’s and Boards step in and take a hard look at how this is impacting shareholder value? Perhaps it can be shown that despite the inefficiencies and fraud there is still a favorable ROI. Or not. Perhaps true transparency in the supply chain is possible. Or not. Perhaps it can be shown that all the money going into the hands of the myriad of players between the advertiser and the consumer – the middle ground – add legitimate value. Or not. Perhaps it can be shown that the money earned by the middle ground is all being used to support legitimate enterprises and causes rather than thieves and terrorists. Or not. Let’s get answers.
Scott , I am served at least 2 FB ads from Section 4 EVERYDAY, disparaging and onboard all in one post, interesting
Let me put an additional spin on this. Traditional media spend supports journalism that can hold governments, corporations and people to account. (Note I say can, not always does). This is essential to a well functioning democracy.
Algorithmic digital advertising and its b***yard child robodialing random cold calling (and bulk junk mail) cannot.
I would ban both. Germany already bans cold calling has done for years
I choose not to have internet installed in my house. I have it on my private phone where my grandkids cant access it when visiting. Thats a parents job to keep children from stuff on internet that can be harmful!! Its not just facebook but the whole internet system. Its up to the individuals to check on facts. You cant blame one source
I’d just assume that digital ads are a motivator for Facebook to push its algorithms that promote rage-inducing content.
speaking of fraudulent ads…
Instead of trying to take a dig at Trump every chance you get you could at least call out people that have even worse histories when it comes to spam and inactive accounts following them. At one point, out of nearly 20 million only about 30% of Obama’s followers were considered legit. https://www.forbes.com/sites/johngreathouse/2012/08/27/celebrities-with-the-most-allegedly-fake-twitter-followers/?sh=4c1d3141ae3b
This has been going on for a long time, well before the likes of Trump took over social media feeds.
Obama didn’t brag all of the time about the number of Twitler or FB followers he had, Trump did so almost daily, only stopped when he was deplatformed. He complained when they removed the bot followers, which they often do, claiming they took away his followers because they hated him. That is why he is mentioned.
Obama didn’t brag all of the time about the number of Twitter or FB followers he had, Trump did so almost daily, only stopped when he was deplatformed. He complained when they removed the bot followers, which they often do, claiming they took away his followers because they hated him. That is why he is mentioned.
Twitter followers, typo.
I don’t know how much this correlates to the inefficiency of the algorithm but how about the fact that AFTER I purchase a “widget” online I get ads for it for weeks or months. I just bought my widget, nobody need two! That is not a characteristic of an intelligent algorithm.
Yapp. I said NO 10 times to an Ad on YT, but they still show it to me almost every day. This must be the power of artificial intelligence 🙂 #lessismore
Great blog—thank you Professor Galloway. I believe that the emerging technology of non-fungible tokens (NFTs) can revolutionize digital advertising again, and with some disruption it will correct a lot of the disaster that advertising has become. I think the professor is prescient and doing a public service to call out the present dangers, hopefully gaining the attention of software disrupters. I hope this siren call creates a movement to change methods that are used across ad-buying. What was a tulip craze for cryptocurrencies in recent years appears now on the verge of setting into competition against each other, hopefully to find a new purpose that actually has social value. Out of this battle I think NFTs will marry with ad-buying technological platforms. High quality NFTs should be exceedingly fair, scalable, economical, and ecological. I think the Hedera consortium that is spreading Hashgraph technology is going to have an outsized positive impact on advertising. I think that startup companies that are using Hedera platform’s NFTs will step in and disrupt the duopoly (Google and Facebook) from inside-out through algorithmic reinvention that uses Hashgraph. (IMHO the best of all blockchain technology is not blockchain, rather it is Hashgraph.)
I ran a small boutique staffing agency in 2006 and it didn’t take me long to learn that my targetedg oogle ad words budget was being ill spent. I ended up paying way more cost per click to Indian Bots than was ever converted to actual job candidates. I’m gathering things have just gotten worse in 15 years,
FB ads are taxed in Singapore. 7% to be precise. Governments can take their slice of the pie via existing consumption / sales taxes. As a SMB owner, FB and G ads bought directly (rather than thru some ad-tech platform) gives meaningful ROI. No other channel comes close.
I don’t think the beast will be beaten down by any of these measures proposed. Certainly it can and will be fine tuned and adjusted to meet the complaints of industry but it will continue. Maybe it will be transformed or self destruct from its own algorithmic short circuits, or through critical evaluation, but trying to bring it to its knees is a waste of time. If advertizers are already tuned into the system’s flaws then the process has already begun. However the marketing system is still geared to pushing products, brands and shades of differences that we are convinced to buy one over the another. Maybe the problem is in the system itself that makes the shades of products from which we are bound to select, buying some things we don’t really need to deal with problems or situations that are not really either. Rather than tweaking or manipulating or passing laws to try to control the beast maybe the whole system from beginning to end user needs to be overturned. Fat chance. In other words the arguments for making the system honest and virtuous are totally specious and untenable and misguided.
I agree on your general points with programmatic advertising, but to make Facebook & Google’s technologies equivalent to the rest of the programmatic ad industry is hugely uninformed. FB & Google own the most first-party data on the Internet and, alone, have every single component you laid out in the programmatic ad chain. Online data quality sucks, certainly, but FB & Google’s suck less by orders of magnitude compared to the random VC funded ad-tech shops that pop up every 6 months. To put them all in the same category of performance is hopefully uninformed, rather than intentional conflation.
And while I agree there was value in more traditional media buying than algorithmic, your post completely ignores the businesses that now get access to Ad Inventory BECAUSE they don’t have to go through a Don Draper. Just ask any small business on FB or Google who saw their sales pick up when they ran a local Ad. Not every SMB is successful using FB or Google or other online Ads, but years ago they wouldn’t have meaningful access to ANY ads quite literally because of all the Don Draper gatekeepers.
Again – agree with your general premise that such automation at scale is rife with imperfections, and we should all take an honest look in the mirror about what to do with surveillance capitalism, but the brush you used to paint the picture in this post was far too broad and lacking the nuance you’re usually so excellent at.
Bravo!
yep “but to make Facebook & Google’s technologies equivalent to the rest of the programmatic ad industry is hugely uninformed”
From memory, people spend about 40% of their time on social media, but advertisers spend 70% of their revenue there, because “that’s what everyone does”. I’m not convinced that’s a smart idea…
great article
How does anybody not notice this anecdotally? I have imbued my soul into my cell phone and Facebook. It knows everywhere I have been, everything I believe, everything I am interested in, everybody I know, all my web searches, and it recommends the most useless crap ever. If it doesn’t just advertise for things I have literally just bought, it will give me ads for things that are utterly irrelevent in my life. Facebook and Google’s algorithms suck.
and yet Section 4 advertises on Facebook.. oh the irony.
Fantastic, humorous, candid & very informative.. look forward to future articles
I hear your points about fraud, but they just don’t reflect the reality that I experience – I generally find digital ads to be very well targeted to me, and I frequently buy the products too
I suppose this means I have too much personal data there
I am regularly targeted by ads for male ‘enhancement’ products, spike heeled shoes, and ‘beautiful Russian women.’ Not sure who the algorithms think I am, but in reality I am an aging hippie (female) who has never worn high heels, let alone shopped for pseudo-Viagra or Russian women. Their sheer incompetence gives me hope.
Does anyone else use the devices you get these ads on? Or share your chrome browser profile?
Scott. You solution is more ‘regulation’. But, by who? The government is inept at policing even laws like antitrust, how would they POSSIBLY create state-of-the-art algorithms to deal with advertising? It would end up being just another Pelosi/Warren shakedown of business. Of course these shakedowns can be avoided with generous campaign funding.
Second, how can our inept government tax the evil algorithms you discuss? Again, they cannot, as the agencies and smart companies would develop workarounds nearly instantly. Smartest people are on the other side of the government and the Civil Service ‘diversity’ hires.
The solution is simple, don’t click through on ads, and minimize time on social media outlets. Not participating in the game you describe kills it quickly. A side affect is that you will be a happier person.
I have worked in the advertising industry for over 16 years, 4 of them at Google. It was traumatizing. Based on that experience, I see the need for a bit of clarification:
1. There is an undeniable amount of fraud in all the steps of digital media buying. It is easy to cheat due to the lack of efficient verification platforms and because most clients don’t understand how digital media works.
2. With a bit of effort, the right tools, and a lot of testing, it is possible to minimize the impact of the holes in the system. You will never be 100% fraud-free, but you can develop a healthy ROI.
3. Google and Facebook are doorkeepers of revenue for a lot of publishers and content creators. Because of that, the two companies have way more power than it might seem for people unaware of how the system works.
4. All that power and money has been used by Google and Facebook to play by their own rules. They claim it is not fair to regulate them as publishers since they don’t create the content. To this argument, my reply is: isn’t using powerful algorithms to curate what the user sees way more impactful?
I would argue that algorithms are taxed. DSP fees and usage are charged state sales tax (where applicable, certainly in NY).
I agree with John – it comes down to what you measure. If the ads you place generate enough of the important actions you wish your prospects to take this could generate a positive and acceptable ROAS. Your need to measure the actions that drive revenue for your business and the cost to generate those actions. If the math works out to a positive ROAS, you may be paying more than you thought for the “good” ads … but I’m OK paying more for ads that work, then getting cheap ads that don’t drive revenue. If you measure against the right success criteria, you can still make good decisions buying online media.
I’ve had a couple of different client-side roles where our team had large programmatic ad budgets, and I’ve fallen for many of ad fraud schemes you identify here.
But by 2018 we stopped caring about ad clicks unless they were from someone we knew. We would track what they bought online or offline. Ad fraud is still there, but this methodology led us to spend up to 90% less with more confidence in the results.
The problem for many companies is that they treat digital advertising as if it were broadcast and try to justify the same types of metrics (reach, CPM, etc). The Internet Ad Bureau encourages this and have bought into the fraud ecosystem.
You are absolutely right that many companies could slash their digital ad budgets. There are probably even more companies not spending enough on digital because they don’t know how to measure it.
10 years ago I ran a test Facebook campaign for a farm to table blog. Most visits to my site were under 1/4 second. When I reached our to FB they said that their software had already detected and removed 100s of fake clicks, likely from FarmVille. But they did nothing about what was clearly incentived or fraudulent clicks that I was actually charged for. Is the fastest growing and most valuable company in history built on a con? Your article suggests so. The larger issue is how we got into this world of advertising based on click rates in the first place. No other media has had to perform at this level. Ads get shoehorned into web pages and inserted in feeds. A disruptive and largely ignored experience. A decent clad click rate is.1%. How many even saw the ad?. Meanwhile the :30 TV ad while largely unavoidable at least tries to tell a story. For the viewer there is a clear value exchange between programming viewed and ads shown. Can we hit the reset button on digital?
If a person is involved in a drunken driver situation there would be a report about what had happened if you was not available than you wouldn’t have any information on was going on cause people have parties all over America so you would be missing in what had happened because was not there.
Thanks
Great read!!!!
Regulate all social media platforms as publishers. Many of these problems disappear instantly.
Too much money sloshing around. Any ‘regulation’ coming out of Congress, especially a Pelosi -run one, will have lots of back doors in the legislation. All bought and paid for by campaign financiers.
Love. Thank you. Can you and Kara please start highlighting anyone who is trying to do an alternative platform? Sooooooo many people are ready to make the switch or have just quit FB, etc. Who is making what’s Next? Esp for SMB’s – what could people look fwd to?
Have you tried WT.Social?
You tackle the symptoms – go after the disease!
Let’s outlaw advertising push; only pull advertising is allowed. Assuming that most of our population is on-line, when you search for an item, then the advertisers can unleash their barrage of information, hype & spin, but if you are not actively looking, they are not visible. Get rid of billboards distracting drivers who should have their eyeballs glued to the road and others on it!
There is nothing wrong with advertising; it is just information – good or exaggerated – but we have let it rule our information systems and economy. We use advertising to continuously encourage us to buy things we don’t need in order to sustain growth – obviously a model that cannot be sustained – which has fairly directly led to the environmental crisis we are facing.
Get to the root Scott; don’t fiddle!
John.
Excellent article Scott. The graphics were really helpful to understand this scam. Sorry for the future demise of advertising, but they’ve hoisted their own picard.
Agree. 100%
I remember being at pretentious advertising boondoggles for years and hearing Draper types laughing about the old adage by the Lever Brothers (Unilever founders) that half of ad spend was total waste, the problem being they didn’t know which half. (hardy har har) I was always surprised that this resulted in the continuation of the same shit vs a reassessment of the entire industry and why and how it worked the way it did. Meet the new boss, same as the old boss…
The fallacy is that the ROI on digital spend is far, far superior to what preceded it. How do I know? Follow the money. Businesses wouldn’t continually up their spending if the results weren’t better. The accuracy of the eyeballs is not as important as the results which are drive ad spend. Sorry to burst your bubble.
Speaking as an agency, is this the role of an agency — an independent that isn’t invested in the vertical ecosystem — to be an expert who provides oversight with exposure to a breadth of clients and innately vested in client performance.
As you say, not every brand is Tesla and advertising, as a form, has been around for centuries. I can’t imagine that going away. The product matters, the message matters, and transparency is more critical than ever. Nothing kills a bad product faster than good advertising, but a good product needs awareness of it’s existence.
I’m not all that, but I recognized about 8 years ago, that one thing, and one thing only, needs to be done with Facebook – shut it down.
The most lucid, compact and informative piece I’ve read on this topic.
Totally agree.
Good Article 🙂
Scott, I can’t fix all of Facebook – but I can fix the shitshow that is election and candidate information part. Here is how we do it: we build an online, mobile, and social campaigning platform – aka the “level playing field” we use to all talk about. On this (free to voters) platform, voters get an interactive sample ballot where they can review all the candidates in all the races up and down their ballot. They can also comment, select favorites, endorse, share, etc… you know – all things we can do when buying a car or TV.
Candidates build and maintain their profiles – just like every other social media site. The difference is this site is built for democracy – not advertisers. There are over a million candidates (running for 519K offices) out there looking for a better way too. We charge them $400 per election. That is enough for me to run the platform without any ads or boosts or bots or fake accounts….
Btw: the platform is mostly built – I’d be happy to demo.
Jim
Thank you for this. I have colleagues who say “Oh this outrage is overblown. They’ve been horrible for along time” and I see apologists on Twitter with the same “whatever” sentiments. But it is the continuous revelations, each worse than the next, that perk up my ears. Thanks. I hope to sign up for a Sprint next year!
Yes, 1st to comment ! So excited. TL;DR by the way..