The first of July 2021 my beautiful daughter was born. For the first time in years, I’ve spent months far from the Agency’s daily struggle. Liberated from the incessant rain of mails, WhatApps, Teams messages, briefs and problems to solve we live our days by; I’ve had long hours of baby naps on my arms to read and think.
I drafted a couple of ideas about my field of action, Digital Advertising:
1.- Digital Advertising could be at the gates of a major reputational crisis. It might not look like it now. Prospects in terms of growth and spending are good. However, there are several issues threatening to mine the confidence in “Digital Advertising”.
2.- This crisis can also become the opportunity for a Digital Advertising renaissance. The shady stuff that has been brewing in the dark is finally being brought to light. This allows us to liberate ourselves from it and get a fresh start, taking a better path with all that we know.
This long essay will focus on the first point. It is an attempt to putting many hours of reading plus a lot of disperse thinking in order so I you can take it in a way that makes sense. But that doesn’t mean that the topics we’ll cover are at the same level or share the same nature. This is a messy cocktail of stuff that -according to my personal point of view- could very dangerously damage the credibility of the whole Digital Advertising practice.
If you find in you the patience to endure through the whole thing, I hope it will give you in return a lot of insight and brainwork (besides a lot of great reads from people way more interesting than me) to get an update and maybe refresh some opinions about the Digital Adland.
I can’t stress this enough: This is 100% a personal opinion, that doesn’t speak or apply for the agency where I work neither the clients I work with.
Enjoy the ride.
The promise of Digital Advertising
I have been a Digital Strategist for over ten years now, and the title has never been less cool and enchanting.
When I started working on Online Marketing -around 2009- it was the Promised Land. We believed -and we told our clients- that we would deliver “the right message, to the right person, at the right time”. That we would reach them in more innovative and engaging ways. That we would be bigger than TV. But most importantly, we told them that -by targeting, measuring and optimizing everything- we would unlock new heights of effectiveness, finally being able to break John Wanamakers’ rule (or Lord Leverhulme’s, it appears that its authorship is unclear, if not made up):
“I know half my advertising is wasted, the trouble is I don’t know which half”.
More than ten years have passed and it is obvious that most of those things have not become true. One is: Digital is now bigger than TV.
According to InfoAdex [Link SP], Digital accounted for 44,5% advertising spend in 2020 in Spain, more than ten points above TV (33,6%).
As you can see in the graph below from Statista [Link SP], Search Engines and Social Media combined surpass TV ad spend, and that goes without adding other digital media like Classified Ads, Portals, E-commerce or Influencers.
The US Ad Spend landscape looks very similar, leaning even more heavily towards Digital, that holds the 48% of overall investment, 7 points above TV (31,1%):
Every fortune has an ugly origin.
Chuck Palahniuk – Diary: A novel (2003)
We will see how -in a pretty counter-intuitive fashion- the more money is spent on Digital Media, the less clear it becomes it actually pays off. But before we get into that, there is something I need to lay down: The Advertising business is all about trust.
Advertising and trust
Trust, buy verify.
The whole advertising business is based on major leaps of faith. Despite the thousands of millions of dollars invested every year, there are a lot of things notoriously “unquantified” or “very poorly tracked”.
For example, in Spain, there are 5.720 people meters [Link SP] (the device that registers what TV program are watched). Out of 18.754.000 households in the country, around 98,6% have at least one TV. This means that the statistics about TV watch time are based on a sample around 0,03%. Theoretically, a statistically valid sample, yes. But still 0,03%. And this is after Kantar increased the amount of meters on 2020; until then, there were 4.875.
Other traditional media don’t have more precise tracking systems either. We don’t positively know how many people see ads in billboards, newspapers or magazines. But we trust our estimates. And more than anything, we trust ‘all that’ will work.
One of the alledged perks of Digital Marketing was precisely “tracking”: Finally campaigns you can reliably measure. As we will see, having numbers at your disposal is one thing, but considering them “reliable” is a whole different story.
Trust drives decision-making:
I have worked with a few Marketing Directors and each one of them had pretty solid preconceptions about which channels to activate and to what end. When a CMO doesn’t trust something, it’s complicated to flip that state of mind. Even if you get them to try, they usually resist giving in the amount of money, effort and love it would require to really work, making it less likely to achieve satisfactory results (and that experience becomes another reason to lack faith on that in the future). A lack of trust regarding a certain media will probably lead to waning budgets.
Besides, the Advertising Industry has a inclination to allow assumptions based on “gut feel” to ground decision-making. Often, these assumptions spread despite not having much empirical back up, but they are still hard to dismantle. For example, we are struggling with this notion about “people not watching TV anymore”, in spite of the evidence pointing in the opposite direction: On 2020 all the records of TV watchtime were beaten [Link SP].
“US TV viewing increased 16% over the last 10 years to 2010.”
“IPA Databank figures suggests that the sales effects on TV advertising are getting bigger, not smaller”.
How not to plan – 66 ways to screw it up
Les Binet & Sarah Carter (2018)
“I’m not upset that you lied to me, I’m upset that from now on I can’t believe you.”
So, where could this credibility crisis come from? Why is “Digital” losing trust? Well, we will mention a lot of things. Most of them are connected. None of them by itself would suffice to change the foreseen media landscape. But merged altogether, they could add up into a new perception from Digital Advertising being shady, inaccurate, low performing or simply a scam. Untrustworthy. And as we said, distrust will likely lead to shinkring budgets.
Facebook is an absolute shitshow
Building a mission and building a business go hand-in-hand.
I am assuming you know about the Facebook Papers, but in the case you don’t: hundreds of documents from Facebook were leaked revealing all kinds of perverse stuff going on inside the offices from Menlo Park. It goes from covering internal reports proving that Instagram is making teens feel terrible about themselves to spreading anti-vaccines misinformation to keep people more engaged.
It’s an unprecedented reputational armaggedon, even for a company that has already been through some serious issues, like the Cambridge Analytica and the Flawed Video Ad Metrics scandals back in 2016.
As soon as the scandal started to emerge, a few leaders from Ad Industry rushed to announce that the infamies being unfolded were unlikely to alter their relationship with Facebook. You can find the exact excuses here [EN] (shame on us).
However, for our clients -Brands- this could be approached differently. I believe that –in the times of Purpose– some Brands could feel discomfortable funding one of the most publicly evil companies of the world. It happened last year, in the #StopHateForProfit Boycotts [EN], and the evidence against Facebook’s malpractices back then was light years from what have been disclosed now. Boycotts could happen again, at a bigger scale. (I have to say that we are already in December 2021 and I have no news from any of this happening, but who knows).
It’s not only about principles, it’s your money too
Besides, one of the things that also popped out in the Papers hits straight to advertisers’ pockets. The Single User Multiple Accounts issue, or SUMA for short.
You can read the whole story here (and I encourage you to):
Long story short: An internal presentation in Facebook revealed the existence of a lot of “SUMA” -profiles that belong to existing users-.
Facebook praises to be a “real identity platform”. Unlike other sites that allow users to sign in anonymously and host as many accounts as you want (like Twitter), when Facebook share their Reach metrics, they want advertisers to expect (quite naively) that those numbers reflect “actual people”.
However, the examination of around 5.000 recent sign-ups shown that the percentage of accounts created by existing users oscillated between 32% and 56%, and the system responsible for spotting this problem tends to undercount them.
Facebook has argued that this doesn’t pose a problem for advertisers, since they use the platform “for their desired results”, not to chase mere Reach numbers. That’s obviously not true.
1/ First, because when advertisers stablish how much money are they going to spend on Facebook, they are counting on flawed numbers of coverage.
2/ Second, because maybe (and this is quite a big maybe) these SUMA are not affecting campaigns that target user responses (like Clicks or Downloads), but they are definitely eating some of the Reach&Frequency campaigns’ budgets.
This is not even a conclusion of mine: it’s stated literally in that internal 2018 document. They said that R&F Campaigns aimed at audiences larger than 10M users might take a hit of weekly reduction that could go from 2,2% to 5% or even 10%.
What really blew my mind was that Facebook answered that “well, R&F campaigns are a small share of our revenue”.
Imagine you found out that one every ten tomato soup cans you buy on the supermarket comes empty. And when you complained to the manager, he answered you: “Well, we don’t sell much tomato soup anyway”.
And we are talking about SUMA: Meaning “duplicated accounts”. I have duplicated accounts on Facebook, but when I use any of them, I am still a human being. There’s no mention in the article about how many of these “duplicated accounts” could be actually “Bot accounts”. One thing is counting people twice, counting people who doesn’t exist is very different.
Just for fun, another memo from Facebook showed that the “active users on their 20s” in the platform often exceeded the total population of Americans that age at the time. Who said Facebook was for grandpas?
It’s very hard to take Facebook’s numbers seriously
It’s not the first time Facebook have had problems with fake accounts, duplicated accounts or whatever you want to call it: accounts that are not worth paying for.
By November 2019, they shut down 5,4Bn Fake Accounts that year (just on their main platform, Facebook). On 2018, they removed 3,3Bn.
And those are the fake accounts they found. It will take you two minutes to find some yourself. Go to any celebrity page (like Cristiano Ronaldo’s) and check out the comments for suspicious individuals.
You will find endearing guys like “Bicho Hermano”, whose entire activity is public and consists on sharing stuff about Cristiano Ronaldo.
If you take a quick look at his more than 3.000 friends, you will find other guys like: Toto Futbolero (who works at “Football”), Andrea Aymimadrelbicho, Bicho Guapo, Bicho Humilde, BichoGod Serresiete, Bicho Good, ElBicho BV, El Bicho Siu and Serresiete Humilde.
Those are in the first scroll of “Friends”.
Remember Facebook praised themselves to be a “True Identity Platform”? Come on.
This is not an exclusive problem of Facebook, it also affects other properties like Instagram.
A recent study in the UK put numbers behind a well-known yet uncomfortable truth about Instagram: There are many fake accounts. Fake accounts follow all kinds of stuff to disguise their activity. That also corrupts Brands’ numbers.
In fact, the study shown that “Brands” were the category with a higher percentage of followers likely to be fake (25%).
It didn’t seem to me that the sample neither the methodology of this study were solid enough to take this average for a fact. However, it does show that the %s of suspicious accounts for brands with many followers are big enough to be taken into consideration.
With everything we know, taking Facebook’s numbers seriously is beyond naive. And “numbers” is the only thing they have left, since they have no credibility and -as we will see later- not much creativity either.
Facebook’s damage goes beyond Facebook
I believe Facebook’s reputational crisis have a negative impact in the whole Online Marketing world.
Not only because of the millions dollars we put in their ads, not because the millions of users we reach daily through their properties, not even because we can’t imagine a media plan without them anymore. It’s more of a philosophical matter. Facebook has been one of the major players to define how we conceive Online Marketing. It shaped the way we understand Social Networks. It’s one of the “Funding Fathers” of Digital Advertising. Their misbehavior discredits all Social Media.
One of the things that enriched Facebook the most have been a growing fixation for Performance Marketing. This too has started to rise a few eyebrows.
Performance Marketing is under suspicion
Performance Marketing has been steadily growing for years. I think it is even fair to say that its growth reflects how many Brands have pivoted the way they advertise during the last few years.
Keeping it super simple, the way Performance Marketing works is:
1/ Find users who are interested in buying your product.
2/ Ruthlessly narrow-target these users with cookies, remarketing, Search Ads and so.
3/ Push them to conversion with direct, straight-to-the-point, rational messages. Use free sampling or price promotion if necessary to be sure you close the deal.
Within Performance Marketing I consider any of its manifestations: Social Media, Search, Programmatic, whatever. Having said this, what we will discuss would admit a lot of debate around the specifics in each field. Please, keep an open mind and understand that we are talking in broad terms.
This type of marketing is typically very efficient, high ROI and with a high Conversion Rates. Sounds great, doesn’t it? Who would dare to question such fantastic deal?
Despite how good it looks, evidence stubbornly shows this kind of marketing alone is not enough to generate growth and profits (specially long-term growth and profits).
Les Binet, -Head of Effectiveness at DDB Adam&Eve and one of the world’s most renowned voices on advertising effectiveness-, explains this in depth but still in very easy-to-follow terms:
Every time I share a link like this, with its author and Title, is my way of recommending you Bookmark it and save it for later. In order to keep you reading, let me summarize the most crucial points:
Binet differentiates between Brand Building and Sales Activation (where Performance Marketing belongs). Sales Activation does not create demand or interest for a product or brand. It only “milks” an existing interest. Pushes this interest into sales.
At the beginning, you will see a spike in conversions. However, if you only do this kind of marketing, at a given time you will lose Brand Awareness, you will lose Mental Availability and the pool of buyers that you can push into conversion will become smaller and smaller.
Besides, Performance Marketing usually has a more rational approach, which means it involves a more price-sensitive decision making. This is also why is very common to include price promotions in these campaigns. That means less profit.
Binet leaves a door opened for Sales Activation, asserting that it should take between 25%-40% out of our budget -depending on the brand and the category-.
Rand Fishkin -Co-Founder of MOZ SEO and SparkToro, a very respected author in the world of Digital Marketing- poses an even more daring hypothesis for Performance Marketing:
What if Performance Marketing is not really “selling anything” per se, what if is only taking credit for sales that were about to happen anyway?
To understand his arguments, we need to get familiar with the concepts of Attribution and Incrementality:
1/ Attribution are the rules stablished to decide which media gets credit for a conversion.
Let’s check, for example, Facebook’s default Attribution Model [EN]:
Facebook selects a last-touch model with a 1-day impression and 28-day click window. For example, if you were to select purchase as your conversion, and apply this default attribution model and attribution window, your reporting will reflect purchases that can be attributed by Facebook to the last ad click that happened within 28 days prior to purchase or the last ad impression that occurred within one day of purchase, whichever happened last.
So… If you clicked a Facebook Ad into an e-commerce, Facebook would attribute to that campaign the credit for any purchase that happened in the following 28 days.
No matter how many times had you been to that page before.
No matter if you bought the products of the page where you landed.
No matter if you bounced inmediately after clicking.
And no matter how many times you went back to that page -and from where- in the following days.
It reminds me of Enrique’s joke about how decisive was the pass he gave Maradona before he dribbled half the English squad and scored the most famous goal in history.
2/ Incremental sales are those being provoked by a marketing action. Sales that wouldn’t have happened if that specific action hadn’t taken place.
What Rand arguments is that the ad tech giants (Google, Facebook, Amazon…) know enough about their users to predict what are they about to buy, so it’s relatively easy for them to place ads of those products in front of them before that purchase happens. That allows them to get massive attribution credit, but it doesn’t mean they are actually driving any incremental sales.
Was that ad really decisive for you to to buy that product, or was it a shortcut to a purchase you were going to make anyway?
At first, you could think: “Well… leave the campaigns running anyway, it won’t hurt, right? If it’s selling, it’s selling. Better than taking the risk of not selling, right?”.
Technically it does hurt you. If you are investing millions (thousands, hundreds, whatever) of dollars on “driving sales” that were going to happen anyway, you are eating a good chunk of your profits. Money you could invest on other stuff that actually generates incremental sales.
Ok… but, all this was always like this, and Performance Marketing has been massively growing for years. We can say it has changed the way we approach advertising at all, so… Why are we questioning it now? Where are these concerns coming from?
Something happened, but nothing happened
Well, something happened. Or, more precisely, nothing happened: For different reasons and at different times, a few giant companies cut Performance Marketing expenses… and they didn’t see any significant drop in their sales results.
I am going to mention three famous cases, but if you do a little digging, you will find other examples:
1/ Airbnb dropped 541M$ in online bidding and search marketing, and they kept 95% of the users they had on the previous year, getting unpaid or direct traffic. They have stated never to return to the same levels of investment on Performance Marketing. Their focus from now on will be almost exclusively Brand Building and PR. You can read the details here:
2/ In 2017, P&G cut 200M$ out of its digital performance campaigns and reinvested them in other media (including TV, audio, and e-commerce). They revealed it helped them remove 20% of ineffective marketing and their reach increased by 10%.
If you want to know more, read on:
3/ Uber turned off 2/3 out of their annual advertising budget -around 100M$– and they saw no change in the effectiveness of their advertising campaigns.
You have the full story here:
What really catches the eye about Uber’s case:
“What we saw is a lot of installs we thought came through paid channels suddenly came through organic. A big flip flop there, but the total number didn’t change.” (…)
“How can we be the victims of fraud? We don’t pay on views or clicks, we only pay when there’s an actual rider who takes a trip in a real car with a real driver and pays with a credit card. And so we realised this was ‘welcome to attribution fraud’.”
If you were thinking you were safe from this “Performance Marketing Bias” because you only work optimizing your campaigns on “conversions”, you could still be wrong. Conversions could be happening anyway; you might simply be paying for them again.
Again, because all these downloads, sales, registers…, are probably not raining out of the clear sky. There is a work of Brand Building behind them. People does not download Uber because they’ve seen an Uber banner. They download Uber because they know Uber. They know what it is, they know what it does, and they know it’s reliable.
Some brands paused digital ad spending in 2020, due to the pandemic. If they saw no change in business outcomes, they can be very selective in what spending they turn back on, because those expenses were not driving incremental business anyway.
Dr. Augustine Fou (Ad Fraud Researcher) – Forbes
Considering what we know about Ad Fraud and Programmatic (the next topics we are going to cover), it shouldn’t be hard to believe that the effects of Performance Marketing are at least overrated.
However, really testing Performance Marketing is a hard step to take. First, it involves questioning a lot of settled pre-conceptions in the industry (narrow targeting, avoiding wastage… but we will talk about all this at some other moment) and -as said at the beginning- stablished notions are hard to dismantle.
Second, it is also risky: What if this campaigns are actually driving sales?
Also, if you switch off and sales stay pretty much the same, there might be a lot of uncomfortable questions: Why were we doing that?
Besides, reallocating Performance Marketing’s budgets is not a painless process either. It might be less effective, but it’s definitely easier to chase some clicks on Google jfand Facebook than consistently building a brand.
I should clarify that, despite agreeing with many of Rand’s arguments, I’m way more aligned with Binet’s vision. I don’t believe Performance Marketing is utterly useless.
I am pretty sure it has an interesting role to play in many cases. The problem is that the industry have been mistakenly assuming that Performance Marketing was the “new way to sell“.
In 2019, a survey from the WFA showed clearly that the industry’s sentiment was that Performance Marketing was really the main driver of effectiveness:
This led to important -and ineffective- switches in media investment. Now, with all these cases from companies cutting crazy amounts of budget on Performance Marketing without getting hurt, many CMOs might start feeling like it is the Emperor’s Dress they are wearing.
Will there be a Performance Marketing Exodus?
Common sense suggests that hype should be going down, but forecasts indicate the opposite. Growth is expected. And nothing is expected to grow faster than Programmatic. Which could aggravate the trouble, since Programmatic is probably the worst kind of Performance Marketing.
Programmatic is pretty bad
Programmatic is based on a very simple principle: Don’t buy spaces -like a magazine page or a blog header- buy impressions for those who you want to reach, wherever they are. Or as Bob Hoffman puts it:
Ad tech’s value proposition is this: we will find you the highest quality eyeballs at the cheapest possible locations.
Bad Men – 2017
It’s a different approach from Contextual Advertising, that goes like: My product is for mums, so I’m going to advertise in All About Newborn Babies Magazine: This way, all the people who see my ad are going to be mums.
Programmatic Advertising goes like: “Mums are mums ‘all day’. When they read their Magazine about newborn babies, but also when they watch the news, follow a recipe or glance over a celebrity’s latest affair. We are going to track these mums and show them your ad when it’s cheaper and more appropriate”.
First of all, this premise implies accepting that the Context where you see and ad doesn’t play a role in its effectiveness.
“I will equally trust this banner about an MBA wether I see it in the Wall Street Journal or in Kittens Gone Crazy.net”.
It’s not only a matter of “how related is the site I’m visiting to the ad I’m seeing”. The prestige of a site plays a role as an endorser for the brands advertised in it as well. Poor quality sites can downgrade the ideas attached to a Brand.
But, being fair, Programmatic does hold a lot of advantages towards a more traditional model advertising (dealing directly with media). Let’s check them out real quick:
I/ Reach: Since you are not limited to the media you can deal with, you can expand your reach as far as your budget allows you to.
II/ Some categories are hard to “contextualize”: The example I shared about mums is not applicable to every category. For example, I like comic books, but I don’t usually browse sites “about comic books”. It would be hard to choose which media you should advertise on to get to someone like me. With Cookie Based profiling is supposedly quick and easy to make this match.
III/ Scale: Programmatic is easy to scale. Setting new Segments, new Markets, new Products it’s only a few clicks away. Imagine if you wanted to enter a new market and you had to research which are the most adequate websites to advertise, contact them, study their numbers, make a deal with each one of them… It would take months of work. Programmatic allows you to do that almost instantly.
IV/ Flexibility and risk: If you do happen to reach an agreement with a media partner and the things are not going too well, you can’t not break it overnight. You will probably have certain obligations. With Programmatic, you can pause or adjust your campaigns anytime.
V/ Reporting integration: If you were dealing with different media partners, you would need to integrate their independent reports by yourself. Clearly more painful and messy than dealing with an only source that aggregates all the data.
VI/ Price: Programmatic is cheaper. Way cheaper.
Well, I think I gave a fair share of credit to Programmatic. These reasons explain why this advertising model has been the fastest growing during the last years, and it is foreseen to keep steadily growing during the following ones.
In fact, I read somewhere that “Digital Ad Spent” is not actually growing “out of Programmatic”.
But its advantages can’t blind us to the fact that Programmatic has morphed into a monstruous and incredibly complex ecosystem, becoming utterly incomprehensible. An intricate net of data collecting companies, data management platforms (DMPs), data serving platforms (DSPs), middle-men weird tech players, agencies, and media networks. I don’t think anybody knows with certainty what data are they buying, where does it come from and who are they buying it to. This is how this Pandemonium looks:
With the impunity this opaque merge provides, bad stuff is happening. Stuff that shouldn’t carry on being ignored:
1/ Targeting is very, VERY inaccurate
“When academics studied the targeting parameters purchased from ad tech data brokers, they found that even a single parameter – gender – was only accurate 42% of the time.
It shows that when two targeting parameters are used for targeting ads — gender + one selected age range — the accuracy drops to 24%, on average. The range of accuracy among different vendors varied from a low of 12.9% accuracy to a high of 32.3% accuracy.”
Dr Augustine Fou
Here’s the source article:
It’s really interesting. Everything he writes is interesting.
Main point being very clearly stated: Programmatic targeting is really bad. And “targeting” is the core-idea behind its very existence. All the advantages mentioned before are fair game if we trust to be reaching -at least to a decent degree- those “eyeballs” we were seeking. If we are hitting anybody’s eyeballs, or not even human eyeballs, the whole think doesn’t make much sense.
Why is Programmatic targeting so bad? First, because the users’ profiling is inferred by analyzing browsing patterns, and that is way more difficult than having somebody first-handly filling in information in a profile-based network (like Google or Facebook). Second, of course, because of ad fraud.
2/ Programmatic is fueling ad fraud
Of course, the great problem of Digital Advertising -but specially Programmatic- is Ad Fraud.
Ad Fraud affects everyone, including the most prestigious publishers, but there is not a better breeding ground for these scam to flourish than this intricated and uncontroled net of traffick bidding, with advertisers buying traffic completely blind to where are they actually being shown or to whom.
“When a small business owner doing Facebook Advertising turned off the Facebook Audience Network, he saw the number of ad impressions plummet by about 90%. He also saw the number of clicks drop dramatically and the effective CPM prices go up. But the sales of his music and merchandise went back up.”
Dr Augustine Fou – When brands stop spending on Digital Advertising nothing happened. Why?
If half the impressions we are kicking are seen by Bots and the other half are so off-target that it would be more precise to stadistically aim to everyone, no wonder companies are cutting off budgets without getting any harm.
Ad fraud is a major problem
According to the World Federation of Advertisers by 2025 ad fraud may be the second source of criminal income in the planet, after drug trafficking.
Bad Men – 2017
I recommend you to read this horrifying yet enlightening article:
Ad Fraud cost 4,5M$ every hour
The most terrible thing about Ad Fraud is that we can’t really know how deep is the rabbit hole. You can’t find any reliable quantification, because whatever we find, we have to admit that it could only be the tip of the iceberg.
The ANA/WhiteOps study discovered that several tactics publishers and advertisers believe are effective in preventing bots are, in fact, mostly ineffective: “Bots faked all of the engagement and viewability metrics we measured.”
It is an extraordinarily difficult problem to tackle. The first and easiest step to take would be stop buying anonymous traffic in sites you don’t control. In other words, stop fueling Programmatic. Or at least, start taking a little control. I’m not an expert on the media buying side, but I assume you can add layers of safety by priorizing white-listed sites and optimizing on certain user-responses.
Nearly 25% of all video ad impressions are fraudulent (viewed by machines, not humans). More than 10% of all digital display ad impressions are fraudulent.
(Beyond the obvious) there’s not a lot of incentive on unveiling that one is being victim of Ad Fraud. Nobody wants to find out it’s their money that is getting stolen and their clicks that are worthless.
“As far as agencies and other marketing entities are concerned, fraud seems to add to their revenue stream.” (…) “There are two groups of people getting royally screwed by ad fraud. First is asleep-at-the-wheel advertisers. (…) You would think that with all the recent headlines about ad fraud they’d start connecting the dots, but they always think it’s the other guy who’s getting screwed”.
Bad Men – 2017
But an industry that wants to get taken seriously can’t keep looking the other way about a problem of these proportions. If we can’t start putting reliable firewalls to prevent our ad spend being stolen by Bots, eventually the only way to solve the problem will seem to be simply to stop buying traffic, and that will leave a lot of people out of business.
I could leave it here. But there is something else I would like to take off my chest. It’s a more personal reckoning, but it counts too.
I believe that all the problems we laid out could be softened by the inmense power of great ad work. But we are not getting a lot of that on Digital Media, are we?
Creativity: Constraints and ‘mass production’
I will focus this bit on Social Media, but the core ideas could be applied to Display and other forms of Digital Advertising.
If you glance at the Social feeds from a few brands in almost any category, you will see that most of what’s published does not stand out.
It is frustrating, because “Digital” is the most rich, interactive, and flexible media there is.
No other support allow to bend the rules and build new scenarios of communication as digital media. But that is precisely what almost never happens. It’s actually quite the opposite: Most of what gets done is predictable, bland and uniform.
Why? From my point of view, two main reasons:
1/ Playing by their rules
As I briefly mentioned at the beginning, this last decade started with Digital as the Promised Land for communication and advertising. In this context, Tech Giants have been looked at -not only as the agents that would allow us to connect with our customers- but as semi-prophetic entities that were shaping the Internet and leading the path for new ways of building success.
We became so obsessed trying to understand how their platforms worked, how could we make the most out of them… that we gobbled down any advice they gave us. And we seem to have forgotten that they have their own agenda.
I am going to use Facebook as an example, because I believe that they are slightly worse than Google or other Social Media at this, but it could apply to them too.
When people from Facebook tell you:
“Users don’t pay attention to long videos anymore; you need to make really quick snackable pills.”
They are actually telling you:
“Our users sistematically skip advertising. They do everything they can not to see it. Despite you were sold on the idea of super relevant targeting, that you were reaching just that people interested on your product and so on, they don’t want to see your ads at all. So make it quick.”
First it was 10 seconds videos. Then 6 seconds. Now is 3 seconds. Soon it will be “can you tell the main thing in half a second? Because that’s the way how people consume media nowadays”.
You are telling me that folks can watch hours of streaming on Twitch or devour a Netflix’s season in three days, but their attention span will not allow them to keep their focus for over three seconds when it’s me who’s talking?
I am sure you have heard about the “shrinkage of human attention span being down to 8 seconds, less than a goldfish“. According to this article, both things are a myth:
Neither our ability to stay attentive is naturally lessening, neither goldfishes lack memory.
People almost never want to watch ads. The difference is that skipping them on Facebook is really easy. You just move your thumb another three centimeters.
But they have been very smart to turn a problem they have –offering Brands a very harsh context to deliver their ads- into restrictions we need to deal with. And we are having it for breakfast.
Make it shorter, make it quicker, put giant short integrated texts, make it vertical, drop the audio. Put your logo at the beginning, this way if they scroll it so fast it can only start to autoplay, at least you will get a brand impression.
Instead of making briefs with insights and ideas, we are filling them with rules, format specs, guidelines and executional checklists. It’s hard to get any bright work out of that.
Supposedly, Facebook provides this guidance to our advantage. They use the data at their disposal to pull learnings about what works best and they share advise to help us maximize our campaigns’ results.
Agencies and clients -immersed in this tunnel-vision about “optimizing and being cost-efficient”- rarely dare questioning these suggestions. Nobody wants to give up any tricks. But, paradoxically, Facebook is sharing the same tips for everybody, so everyone is doing the same and nothing stands out.
Besides, I have a hard time looking at some of these recommendations and not seeing patterns of poor user experience disguised as “tips”.
What follows are examples that come straight from the Facebook’s Creative Guidance Navigator:
“Drop the audio to increase your VTR”:
I can only makes sense of this by assuming that the audio popping in autoplay is so annoying that only makes people scroll faster. What other reason could it be to justify that audio-off content has a longer VTR? Does anybody who willingly watches a video not tend to enjoy it more when it comes with audio? Besides, they talk about an average 12% lift in VTR. How long is 12% in a platform where videos usually are 10-6 second long? (Yes, between less than a second and 1.2s).
“Place a logo at the center of your stories”:
Yeah, you place a big-ass logo there in the middle, sister. This tip is under the category of “Building Brand”, by the way.
Don’t mistake my words: I am not saying we should ignor the context where our ads are displayed. That would be absurd. It’s crucial to understand how Facebook works, and how people behave on Facebook; but knowing the rules should work to our advantage, not against it.
What needs to be adapted is not the idea to the platform, but the other way around.
If I want people to watch a 30 seconds video and I know I am not going to get that on Facebook, I better buy this traffic to get out of Facebook and offer them a compelling experience somewhere else (in my website, for example). I will lose a lot of people in the way, sure, but it will be better than chopping 30 sec into an incomprehensible 6 sec version that will be getting a 40% VTR anyway.
That, I am sure, is highly inefficient. But remember what we discussed earlier about super efficient campaigns that were not driving any sales. You can get a lot of very efficient and completely useless views, because the impact of a 30 sec ad that a person watches willingly does not compare to the partial view of a 6 sec video that a person half scrolls. Not matter how much money you are saving, what you are buying is worthless.
Am I saying you should NEVER do a 6 video ad? Of course not. If you can say what you need to say in 6 seconds, by all means, do it. If you can do it 3 seconds, all the better. What I am saying is that meeting “Facebook’s standarts” -or any other platform’s- will not necessarily meet your needs.
2/ “Quantity” rules over “quality”
Second reason why there’s a lot of boring stuff poluting brand’s feeds out there: We are collectively possessed by the “Content Plan hysteria“.
Somehow, the idea that brands need to keep a constant content stream in every social platform was settled, and Social Media turned into a nightmare.
Every month, we look terrified at Calendar’s blank spaces wondering… how are we going to fill all this?
It’s crazy how much are we forcing brands to talk. Why?Advertising never had the same role than content creators. And the truth is that most brands simply don’t have so many things to say. However, nobody seems to be willing to give up on the publishing cadence.
Since discussing “quantity” is out the question, this inevitably ends up tackling “quality”:
1/ Poor ideas: We are so desperately looking for things to say that we all end up finding the same solutions: Tips, How To’s, Challenges, Giveaways, Dayketings, Quotes, User-Generated-Content, Memes, Trends. And publishing a lot of posts about our products too, of course.
2/ Poor craft: Very few brands can afford Creative (Strategic, Account, Production) Teams that can deliver 10, 15 or 20 great contents per month… for Instagram. What about YouTube, Facebook, Pinterest, Twitter? Forget about it. We need to make things quick, easy, cheap. That comes at the cost of craft. Quick, easy, cheap things can be done, but they are hardly distinctive. They don’t usually stand out.
3/ Poor Reach: When you are managing big numbers in terms of posts, it also becomes unlikely to have enough money to provide high Reach for all of them. Many end up being Organic or having a very discrete distribution.
These three issues intertwine, aggravating one another.
A lot of work, frustration and suffering out there to achieve very little. If we invested all that talent, effort, time and money in a few good ideas, I am positive that results would skyrocket. But to do that, we need to give up the idea of constantly publishing.
“What are we going to do? Are we going to stop posting for two months?”. Yes. Why not? Who would notice?
There are very few brands whose contents some people actually expect and those are not the brands that should quit. Regarding the other great majority, they could go quiet tomorrow and no-one would know.
Forget your marketer-self and think as a “normal person”. From the brands you follow, is there any of them you would realize have stopped posting if they did?
Nobody is going to make the headlines with yet another Content Plan. The only brands really getting noticed by their work on Social Media are precisely those with a disruptive approach [but I’m going to save this for a potential second chapter of this essay].
“Give me your Likes, Bicho!”
One of the reasons behind this “Content Plan Hysteria” is that Facebook also sold us on the -very questionable- idea that Engagement is a relevant metric for a brand’s health in Social Media. I talked a little about Engagement Metrics on Social Media in this post, in case you want to take a look:
Problems analyzing Social Media
I have read a lot about advertising effectiveness during the last months and I have never ever seen any mention to “Engagement” as a relevant indicator for growth, market share, effectiveness or whatsoever.
Remember Bicho Hermano and his more than 3.000 friends? Are we really going to pay that much attention to anonymous and untraceable Likes or Comments that could be coming from anywhere?
Of course we can use Engagement as a tool to evaluate the reception of certain contents. But I believe that turning Engagement into an “end goal” only benefits the platform where these interactions are taking place. We end up making content that gets engagement, not content that really works towards our bussiness goals.
First of all, congratulations. You are brave, my friend. You made it through.
More than 7.000 words later, I hope to have at least provoked some thoughts. You might agree that the threat looming over the credibility of Digital Advertising is very real, or you can stay skeptical about the impact all these factors could actually have.
Be what may, if you work in the industry, it is also in your hands to change it. Nobody have power to change everything, but we all can drive small improvements in our daily jobs. And the little things add up.
This first chapter took a lot out of me. If you liked it and you would be willing to read the next one, let me know. Write to me in Comments, Tweet me about it, add me in LinkedIn and send me a message with your thoughts. Your feed-back might make the difference between me deciding to leave it here or bringing myself to publish the next one.
Always, thank you for reading.