Man vs. machine: the battle for marketing automation supremacy

Everyone wants their digital marketing to be cheaper or better. But which is better, manual or automated? Let’s find out.

The Matrix, Blade Runner, and 2001: A Space Odyssey are great examples of captivating Man vs. Machine stories. While competition in the digital advertising space might not be as bloody, it’s just as fierce, especially when you’re competing against rivals who seem like they have an advertising budget that could fund a sci-fi movie.

In the digital ad space, the battle between man and machine centers around one big question: Can you successfully automate digital marketing with the help of AI?

 

Can you automate digital marketing with the help of AI?

The short answer is yes. In advertising, robots don’t hook people up to vats to use them for energy. Instead, sound machines, like our advertising robot, can help you to:

  • Automate ad creation
  • Improve the cost-efficiency of your ad bidding
  • Test your copy and creatives to maximize their impact
  • Serve your ads to the right people at the right time

But to answer that question, we need to break it down into smaller sub-questions. There are many ways to look at the broader question, but we prefer to define it by answering these three questions:

  • Are machines intelligent or more intelligent than humans?
  • How do you optimize performance marketing?
  • Which is better suited for the task: man or machine?

Let’s deal with each individually.

Machine Intelligence

First, let’s first start with intelligence. It’s a somewhat controversial topic. Many people think that their work is too complex, contextual, or creative to be taken over or done by a machine. Also, there are many different ways we can define intelligence beyond processing information, such as emotional intelligence.

But the best way that people can wrap their heads around this concept is to look at one of the most famous examples. Magnus Carlson is the best chess player in the world. He’s the highest-rated player ever, so many people in the chess community consider him the best of all time.

However, modern smartphones can beat Carlson at chess. Of course, this is nothing new. In the mid-90s, IBM’s Deep Blue computer beat Garry Kasparov over two sets of six-game matches.

We can all agree that chess requires intelligence. The more intelligent you are, the better your chances of being able to compete at a high level. So, considering this, we can establish that some types of machine intelligence are smarter than humans by a significant amount.

Next, we need to consider whether Magnus Carlsen finds Google Ads too difficult to wrap his head around. Would it be too complex for the five-time World Chess Champion to optimize a Google Ads account?

Or, to take it further, would the people who run and optimize your digital marketing be able to turn their hands to chess and reach the same heights as Carlsen?

We are deliberately provocative here to underline a point: Some domains of human activity are better suited to machines.

Data-driven marketing

Let’s suppose that we know that for some types of intelligence, machines are better suited and measurably smarter than humans. To figure out how to optimize digital marketing, we need to understand performance marketing and whether it suits robots.

Data-driven marketing has been an industry buzzword for a while. At its core, it looks at historical results and evaluates and draws conclusions from the past to optimize future performance. That process is perfectly suited for a machine.

Let’s look at homepage banner ads as an example of what performance marketing can do.

Say you have a website with 1000 daily visits with three banners visible above the fold. Everyone who visits your site sees these three banners. Knowing how to pick what to show on these banners to suit every visitor is hard. People respond to different colors, images, ad copy, etc.

Additionally, If you have a store that sells men’s and women’s clothing and an audience evenly split between men and women, you will be showing the wrong ad to 50% of your audience all the time.

In many cases, static banners can work for the majority of people most of the time. But if you compare a static banner against a personalized dynamic banner, as long as it’s built on top of some kind of evidence, it will outperform the static banner every day of the week.

When you compare smart optimization against manual optimization, machine intelligence will outperform manual work if you give them the same references, frameworks, and data for the evaluation.

You can have tens or hundreds of thousands of search queries on small to medium-sized accounts. All these search queries have some importance. While not all of them are equally valuable, they’re worth thinking about.

And to do that manually, on thousands or tens of thousands of keywords and phrases, is an impossible task. Throw on top demographics, web browser history, and remarketing lists; you have some big decisions to make. Which are most important? Which elements should you consider in relation to each other? There are so many variables to think about that it quickly becomes overwhelming.

Another thing to remember is that when you want to optimize your marketing, the machine must make the decisions. You don’t want a machine that recommends decisions or a framework that needs to be inputted for every campaign or account.

A big part of machine intelligence is letting the robot make the decisions. You need to allow the machines to evaluate which clicks, users, and products are allocated the budget.

Do machines and humans work similarly?

We’ve had many customer questions throughout our time offering performance marketing solutions. For example, when our robot optimizes banner ads, we’ve received emails like “Why are you showing this user this ad?” or “why are you putting a USB cable on the front banner? We know that this doesn’t work.”

This happens because machines and humans work in very different ways. You can’t evaluate a machine the same way you consider a human — at least not based on their process. You can compare them based on the results they generate because both entities are working towards the same KPIs. In short, if the arrows are going up, you can say one or the other is doing a good job. However, their work is vastly different, and expecting machines to act like humans isn’t realistic.

Machines allow you to test a lot more. And it’s necessary to do it because machines don’t come into the process with a priori assumptions or inherited knowledge that come from outside the particular task they were designed to do. For example, a human knows about things that aren’t strictly related to performance marketing which can help them make more informed marketing decisions.

The machine, on the other hand, needs to be tested. You need to teach it either with historical data, which most, if not all, companies do when they are working with automation. But you also need to test things to give the machines a chance to learn what is and is not.

This holds if we look at any type of mechanical development. When we were first designing a plane, we could have created something that would replicate the flapping of a bird’s wings. Instead, we have jet engines. They are more cost-efficient and faster, so comparing the two is unfair.

If you want to take the fastest route from point A to point B, you could build something bipedal that tries to mimic walking or make a motorcycle. A bike could beat Usain Bolt for speed every day of the week. But by the same token, Usain Bolt’s wife wouldn’t want to be married to a motorcycle because he is more than just the task he performs.

Those comparisons get to the heart of the matter. Machines and biological life go after the same goals using different processes. If you are concerned about the final outcome, compare them based on that, not their journey.

Future of digital marketing

If we look toward the future of digital marketing, it will only grow in complexity and sophistication. AI optimizes mechanical tasks, so what we have left is complexity.

Suppose you choose which products to promote via email marketing, on your hero banner, or which keywords to use in your Google Ads account. Those are relatively simple tasks to solve for a computer. What remains are more complex questions like:

  • How does each marketing channel interact?
  • How can I build the best brand?
  • How can I boost customer experience?
  • What happens after each click?

We’ve evaluated thousands of accounts and found that brand is more important than prices, CPC, or click-through rates. That shouldn’t come as a surprise to anyone. All other things being equal, the better brand will win auctions 99% of the time.

People have become a bit lazy in thinking about digital marketing in the last couple of years. A lot of the time, just turning up means you can get conversions, but that’s changing quickly.

Let’s compare Google Ads to a brick-and-mortar shop. You have three types of people walking past:

  • Those who want to come to your shop
  • People who want something you sell and are walking by
  • People who are just walking past.

It’s the same with Google. You have your street, which is Google, and you have different search queries. Some people are searching for your brand, some are looking for something more general, while others are looking for information.

In the same way, we have physical ads, like billboards, flyers, and magazine ads; in the digital world, you have Facebook, Instagram, and influencers. But these days, being visible isn’t enough.

The most important part is the entire customer experience. Good ads and good products can only get you so far. To be successful over the long term, you need excellent customer support, loyalty, and recurring purchases. You need to consider customer acquisition costs, cross-selling, and upselling and how to do that most efficiently. Once the mundane and repetitive tasks are taken care of, you can start looking towards these questions to boost your sales and revenues.

 

Related content

AI has opened the doors to incredible marketing opportunities in marketing and beyond. However, like any innovative technology, it also has the potential for misuse. Brands and agencies need to understand the ethical questions that come with using AI for digital marketing.
Universal Analytics (UA) stopped processing hits on July 1st this year. Now that the sun setting is well underway and GA4 has been in place for a while, it’s time to see how it’s all going for sales and marketing teams. So, let’s have a look.
European consumer confidence dipped to -18.7% in September. That number is better than last September’s lows of almost 30%, but it reveals the hurdle marketing teams need to navigate to drive sales. Here’s how you can take stock of your marketing during these challenging times.

In this article

Stay updated

Get on the (email) list!

Stay updated

Join our newsletter