What Marketers Don’t Understand About Performance Marketing

With Performance Marketing Expert, Dennis Anzalone-Keesee

Dennis Anzalone-Keesee, advertising specialist at Elumynt, shares from his experience the misconceptions marketers have about performance marketing—what it is, and how it works.

What drew you to paid media initially and how has your perspective on performance marketing evolved as you’ve moved into a more strategic space?

Initially, it was purely a matter of financial survival—just a need to better my financial situation! I was working as a night porter in a hotel in Brighton (UK) at the time and used the very long, boring nights of inactivity to study the digital ad platforms, how they worked, and how they related to what I studied at school. It was fascinating, and it was also interesting that a lot of people were actually deploying tactics that didn’t make sense given how the algorithm delivered ads.

When I entered the field professionally, I initially fell into that trap too, using tactics that looked good on paper but, in reality, were not incremental to the revenue of the business advertising. My perspective definitely evolved, but it did so more in relation to how performance marketing is defined and executed rather than the actual reason behind its existence in the first place. There is a lot of confusion in the industry about what performance marketing truly even is. I wrongly assumed practitioners in the industry knew, but I discovered over the years that this is just not the case!

In your own words, how do you define performance marketing today? Where do you think people oversimplify what it actually takes to drive results?

In one word: performative! And I find it ironic that the words performance and performative are so similar. Words are very important. You know—they—mean things! Because of that, I believe marketers need to slow down, take a breath, and look at the meaning of the words they are using.

As a matter of fact, I’d like to do it here.

Let’s look at the meaning of the word “performance.” It can mean essentially two things: it’s either an act you put on for the pleasure of someone else (like when an actor performs a script) or the execution of an action. I think you can already see the flaws when it’s paired with the word marketing.

Let’s take the latter definition for a second and bolt it into the second word: “marketing.” Now, what is marketing? “Marketing” is defined as the business activity that involves finding out what customers want, using that information to design products and services, and selling them effectively. So let’s put the two together and you’ll see how weird even the concept of “performance marketing” starts to be. When you put the definitions together, you can say that performance marketing is either the execution of marketing, or acting marketing. So it’s either just doing marketing or pretending to do it. See the issue?

Performance marketing took a turn to just mean “buying ad space in digital spaces.” That’s it. The word doesn’t describe its meaning, and that creates an insane amount of confusion regarding what someone doing “performance marketing” even does.

Let’s look back at the definition of marketing. Where does the action of buying ads in digital spaces fit in that definition? At the very end. The part that says, “and selling them effectively.” So performance marketing is not even marketing; it’s advertising: the promotion part of marketing, one of the 4 Ps. Which is a big difference! It’s essentially like saying one tree is the whole of planet Earth. A tree is part of the planet, not the whole planet! You can’t say an engine is the car; the engine is just part of what makes the car. Hence why, when asked to define it, I say that performance marketing is performative. It’s mostly an act to please someone, which is usually the people spending the money on ad platforms (the investors, your boss, or perhaps yourself if you own the brand).

In terms of oversimplification, I believe where it happens most is in the actual understanding of the meaning of the words used, like the example above. If you do not know what a word means, you are not going to be able to [utilize it]. You can’t fulfill its intended use. If you call a banana a hammer, and you use it as one, I guarantee you that nail won’t get through the wall!

These words are usually ratios like Return on Ad Spend (ROAS), Marketing Efficiency Ratio (MER), Customer Acquisition Cost (CAC), etc. These metrics are not easy to calculate. I am going to give you a quick example with ROAS. Ask a marketer at random what it means. They will probably tell you something along the lines of “It’s the return you get from your ad spend.” Simple enough. Then ask them what “return” means. That’s where the rubber meets the road, because that “return” is something very specific. It’s the revenue CAUSED by the ad spend. Meaning that if the ad spend hadn’t been deployed, the return wouldn’t have materialized. However, the way it’s reported across businesses rarely accounts for the causality part of the equation.

You’ve worked across eCommerce, affiliate, and agency environments. How have those different models shaped your approach to building and scaling paid media programs?

They definitely gave me a better understanding of the different reasoning behind what people who want to grow their businesses want, or what they think works to grow revenue or profit. That is pretty invaluable, but I’d say it’s more valuable for understanding the clients you are serving better, because it didn’t change my approach much.

You will see that across a variety of different product or service categories, people tend to always want the same things. If I want to buy a car, it’s because I want it or I need it. If I want to buy a bar of chocolate, it’s because I want it or I need it. If I want to buy software, it’s because I want it or I need it. We all share more or less the same brain, and our brains haven’t evolved much in the last 300,000 years. It’s still the same brain wired for the same responses. I mean, look at stress. It is literally a mechanism designed for us to either fight back against the lion trying to eat us or to run away as fast as possible. We are not chased by lions anymore, but did stress disappear?

The mistake I see when people think different product or service categories require wildly different approaches is very common, and it happens because they don’t realize the brains are the same. Some small execution parameters may change here and there, but the fundamentals, which are what marketers need to focus on, and the things that move the needle the most (revenue and profit), never really do.

Efficiency alone is not a metric any marketer should be focusing on without first having determined the effectiveness of a marketing activity.

What are the biggest misconceptions you see from brands when it comes to performance marketing—especially around attribution, optimization, and success metrics?

I will probably open a bit of a can of worms with this question, but it’s an extremely valuable one, so I’d like to go step by step.

Let’s start with attribution. To do that, let’s briefly go back to the definition of marketing: the business activity that involves finding out what customers want, using that information to design products and services, and selling them effectively. Let’s also touch on what “effectively” means: [It’s defined as] in a way that successfully achieves the result. In other words, the outcome of executing marketing is ultimately that more product gets sold compared to if the marketing was never executed. Let’s also assume that by “performance marketing” we mean “advertising,” as that’s what it usually means.

Now, let’s define “attribution” too: It’s the act of saying or thinking that something is the result or work of a particular person or thing. So attribution, in the context of advertising, is the act of trying to say that a product being sold is the result of advertising. That is definitely a noble purpose. However, the way attribution in digital advertising is usually determined (this is how every social media ad platform determines it, how Google determines it, how TV determines it, how every third-party multi-touch attribution tool on Earth determines it, and how, incredibly, even supposed marketing mix modeling tools determine it) is by detecting a touchpoint. This is a moment in time where someone was tracked (via an action on a website detected by a pixel or an in-platform touch like a tap or a view) taking an action, and, at some point after that action, the person bought the product.

Here is where that logic starts breaking in an impossible-to-ignore way. Let’s say you ate curry last night, and after the curry, you had a piece of cake. You go to bed, and the next morning your stomach is completely upside down. What caused your stomach to be upside down? The curry? The cake? An infection that has nothing to do with either the curry or the cake? You see how difficult it can be to determine the outcome of an action when a lot of things are happening at the same time? That is why single and multiple touchpoints are NOT a reliable way to say whether it was that ad that caused the sale and not something entirely unrelated. So, the single biggest misconception in how attribution is usually reported on is that a touchpoint equals causality.

Now, about optimization. Let’s define that to start. What does it even mean to “optimize” something? It’s essentially two things: either making something more effective (like making a car go faster) or making something more efficient (like making a car burn less fuel). Optimization is also limited by the physical reality we are in. We can’t make a car go faster than the speed of light on Earth, no matter how hard we try, for instance. So that’s already a big thing to consider: the fact that you need to have your head anchored in reality, and not try to optimize for things that are impossible in the physical world.

Now, if attribution wants to solve for identifying what caused something to happen, but it doesn’t do it in a way that satisfies causality, then what are you optimizing toward? Something not real. So, the biggest misconception in relation to optimization is that there is no limit to it, and that it’s based on accurate data given by attribution.

Lastly, success metrics. This very much ties back to the previous questions where I explained the issue with marketers not understanding the meaning of the words they say. That illiteracy is also tied to confusing efficiency with effectiveness. Most of what you see in marketing in relation to success metrics is based solely on efficiency. Almost no one talks about effectiveness.

Let me give you an example. Why are cars built? Because they can bring us from point A to point B without walking. That is effectiveness. Now, let’s say my car burns 0.2 gallons per hour to run the engine. That metric (gallons per hour) is efficiency. Now, let’s assume I take the wheels off my car. I turn it on and let it burn all the fuel. I can still measure its efficiency based on how many gallons per hour it burned, but did I go anywhere? No, I had to walk. Measurable efficiency, zero effectiveness.

And here is where it gets even weirder. Usually, a car at idle will consume fewer gallons per hour, so you’d actually have greater efficiency. Still, you are not moving; the car is not working for its intended use. Efficiency alone is not a metric any marketer should be focusing on without first having determined the effectiveness of a marketing activity. Effectiveness is the ability to produce the intended result, which in the context of marketing is producing the outcome of more products being sold due to marketing. ROAS, MER, CAC, and all these ratios people talk about in the context of performance marketing are efficiency metrics that tell you absolutely nothing about the effectiveness of a marketing activity.

So, in the context of success metrics, I’d say the biggest misconception is treating efficiency metrics as signs of effectiveness without doing the work to check whether the marketing activity was actually effective first.

SUBSCRIBE TO OUR FREE MONTHLY NEWSLETTER

When you’re evaluating channels, platforms, or partners, what separates something that looks good on paper from something that actually performs in the real world?

Whether it created a change in the real world! And that change can mean a shift in behavior that precedes sales coming in for a business (like a brand search) or an increase in revenue or profit that would not have happened if those channels, platforms, or partners had not been deployed.

There’s more data available than ever, yet many marketers feel less certain about what’s actually driving outcomes. How do you distinguish between meaningful signals and noise when optimizing campaigns?

You are so right. The first thing I’d do is challenge any marketer to find proof that more data led to more effective advertising. You won’t find any. The handful of studies you can find relate to Click Through Rates (CTRs), and high CTRs have been proven (by a lot of studies now) to have no causal impact on the revenue or profits of a business. If you find any study proving otherwise, please hit me up! I’m happy to be proven wrong.

In terms of meaningful signals versus noise, I’d say three things:

Ensure your ads are delivered and optimized to be delivered to humans. A tool like FouAnalytics, for instance (built by the amazing Dr. Augustine Fou) can tell you that. First of all, if you set it up correctly, it can allow you to see which placements, audiences, CTAs, etc., are getting you real human traffic versus bot traffic. Secondly, it allows you to optimize for more human attentiveness by allowing you to see the taps and scroll patterns of people who land on your ads. It’s a great tool, and I suggest any digital advertiser install Fou ASAP*.

Look to see if your brand searches are going up. This is the easiest proxy to understand whether your advertising activity is creating meaningful awareness about your brand. As explained above, no one accidentally types your brand name into Google. They saw it somewhere. And if you isolate the impact of that activity with incrementality testing, you can clearly see if the cause of the increase was your advertising or not.

Run incrementality tests! This ties back to point two. The most straightforward way to do so is with geo-testing. There are a ton of ways to run them, and I developed a framework myself to simplify planning and measurement for marketers. So if you are confused, hit me up, and I will help you.

How is AI changing the way you approach campaign management and optimization—and where do you still see human judgment playing a critical role?

It’s not really changing it much, and where it is, it’s changing it for the worse by creating more issues—like turning on AI features in digital campaigns automatically. These are features no brand or consumer wants to see, and no marketer in their right mind should deploy them! Broadly, I’d simply say my views on AI align very closely with those of Ed Zitron, Mo Bitar, and Cal Newport. The very fact that we are asking where human judgment still plays a critical role, frankly, is problematic. Shouldn’t human judgment always play a critical role in anything we do?

My hope is for more people to use critical thinking and to question the metrics they see in their dashboards.

How do you see performance marketing evolving over the next few years as privacy regulations tighten and traditional targeting signals become less reliable?

The truth is that I do not know. I suspect we’ll see the continuous growth of new businesses popping up, but the divide between great advertising management and mediocre, AI-driven management will get wider and wider. You will likely have a lot of brands at the mercy of automation, resulting in extremely poor results, failures, and bankruptcies, alongside a lot of brands partnering with amazing, human-led agencies that will continue delivering great work. My hope is for more people to use critical thinking and to question the metrics they see in their dashboards. I hope for more experimental data and less observational data, essentially!

There’s a growing conversation around publishers taking back control of their data and supply paths. How do you see that impacting performance marketing from the buyer side?

I think that would be great and will ultimately ensure that more humans consume ads, thus making profit margins for brands go up, as less advertising budget will be wasted on invalid traffic. If that happens, I can’t wait! It means that more human conversations and relationships can happen.

Because here is the thing: humanity is at the core of everything we do in advertising. Remember, it’s about the needs and wants of people! That is how you effectively sell a product. You meet the needs and wants of a human being. It does not get more human than that. Somehow we forgot along the way, and it is my hope that we will remember again!

If you had to challenge one widely accepted “best practice” in performance marketing today, what would it be and why?

I recently stumbled upon a comment on LinkedIn that read: “Creative quality and output is the media buying these days. There’s so little button clicking left.” I believe that assumption is the one that needs to be challenged.

There is a lot of button clicking left, arguably way more than before. The reason is that, yes, it got way easier to simply launch a campaign without having to touch the settings, but exactly because of that, the strategy behind those choices matters immensely more. It’s easy to launch a campaign, which means it’s easier to burn money. I see it time and time again with brands just trusting that digital ad platforms have their best interests at heart when that is simply not the case. The digital platforms are there to take your money and give you what you want. But many brands do not know what they want, or they do but don’t fully understand how to get there.

I am an aircraft pilot, and you are probably familiar with the concept of autopilot. But what you may not know is that in an emergency, autopilot is the first thing that gets turned off in an aircraft so that a human can manually fly it. Autopilot can literally kill you if left on. The same thing applies to media buying. There are a lot of good uses for autopilot (like mass reach), but also very bad uses (like auto-AI features and auto-bids). It’s wishful thinking to believe that media buying doesn’t require human intervention any longer. Theoretically, it’s a great concept; practically, not so much! In the last two years, I have seen more brands wanting help than not. If automation were truly that effective, wouldn’t we see fewer brands asking for help?

*This is Dennis Anzalone-Keese’s recommendation and does not constitute an endorsement by CreativeScience.io.

Share this on:

• LinkedIn
• X
• Threads
• Reddit
• Mail

About Dennis Anzalone-Keesee