The faded image of marketing is one that’s reminiscent of Mad Men, where teams rely on creative genius and good fortune to score victories. In reality, today’s marketing world more closely resembles Moneyball — the 2011 film about a struggling baseball team that halts a losing streak by using data to inform hiring decisions.
Marketing has become a highly automated domain, where human decision-making is increasingly supplanted by computers that can calculate and act faster than a person ever could. Number-crunching algorithms are responsible for ad placement and delivery. AI systems now write and edit copy, and can even produce bespoke audio adverts for streaming radio and music services. Through data and the adtech ecosystem, marketing teams can understand their target audience unlike ever before.
For marketing leaders, it’s easy to understand the allure of the algorithm. This attraction doesn’t just come down to cost, but also a pressing need for agility and to demonstrate results. As Gartner pointed out earlier this year, only half of US CMOs can prove the value of marketing to the rest of the organization. A system that promises to improve results by using scientific rigor and empirical data is a tempting proposition.
Data, AI, and automation aren’t going anywhere. Nor should they. Many of these tools perform a vital function and act as a force multiplier for marketing teams. And yet, organizations should be wary about outsourcing too much of their thinking to machines. As with all things, they should be used in moderation.
It sounds corny, but a human knows how to craft something that connects to the audience on an intimate, heartfelt level. The ineffable combination of tone, timbre, and tempo that truly resonates. Whereas a bot can tell you what worked previously, a human is prepared to make ambitious guesses based on instinct, observation, and experience. Those guesses often produce winning results. AI is not close to this just yet.
The challenge is to reconcile the human creative with machine-driven analytical processes. When you strike this balance, you get the best of both worlds.
Data Provides Guide Rails For Creativity
Most copywriters are familiar with the daunting feeling of starting a new project. Those agonizing first few minutes where it’s just you and a blank Word document. Not knowing where to start, or how. Half-formed ideas float above your head, all too far to reach. It’s an unpleasant — if not terrifying — sensation.
It’s here where data often shines, giving creatives those important context clues that will help guide their work. When you know the intended audience, for example, it becomes easier to craft language that speaks directly to them. Creatives can look up at the ideas circling the ceiling and ignore the ones that don’t make sense or won’t work, and select the ones that do.
This focus is an often-overlooked benefit of data-driven marketing systems. We emphasize the benefits of demographic insight and accurate delivery, but ignore the fact that data is often a great way to kickstart the creative process.
I started with this point first because I wanted to re-state that marketing teams shouldn’t choose between analytics and human creativity. The best teams use both in a way that balances each’s strengths and weaknesses in order to achieve the best results.
Human-First Problem Solving
If you’ve ever bought an expensive new kitchen gadget, you probably spent the first few weeks using it at every possible opportunity, even when it didn’t quite make sense. To justify the expense of your new Instant Pot, you used it to cook dishes where it would have been easier — if not quicker — just to use a normal griddle pan.
The same is often true within marketing teams. They spend their budget on a new system and start looking for ways to extract every conceivable cent of value, utilizing their latest bauble in ways that might not deliver the desired results. I can imagine AI copywriting tools — like ChatGPT, Grammarly Go, and Jasper — being used this way.
The problem with using technology to solve fundamentally creative problems is that they only produce content that has previously been proven effective. And this means creating work that’s clichéd, tired, and over-optimized.
The ability to detect inauthenticity is almost like a sixth sense. Everyone has it. We all instinctively know when we’re being pandered to, and we’re programmed to reject it. And so, things that worked in the past will soon become ineffective.
This view is shared by legendary adman Rory Sutherland, vice-chairman of Ogilvy, who warned about the potential consequences of automating creative work like a “Fordist dream of multiple copies rolling off a creative line.” The best examples of copywriting, he said, were the product of multiple human voices providing feedback and suggestions to the original author, refining and improving until it reaches total perfection.
That’s not to say data, automation, and AI don’t have roles to play in the creative process, but they should be minimized to suggestions. As Sutherland said, “[AI] is like satnav in a car. Great for directions but you don’t allow it to drive the car.”
Don’t Be Afraid To Color Outside the Lines
Analytics are great for telling you what’s worked previously and helping predict what will work in the future. This is a useful tool for marketers, but it can introduce an element of blindness where they’re so constrained by a winning formula, they lose the ability to experiment.
This blindness becomes a problem when you consider that trends are fleeting, and consumer sentiments can change at the drop of a hat. Just think about clickbait titles — deliberately vague headlines designed to provoke the reader’s curiosity and entice them to click. At one point, they were charming and highly-effective, with sites like Upworthy and Buzzfeed using them to great effect.
But eventually, they got old. Consumers became aware that their attention and curiosity were being manipulated by traffic-hungry sites. Suddenly, “clickbait” became an epithet and sites that used it were regarded with scorn and disdain.
That was an extreme example, but you get the idea. The reality is that if you rely on the same old style and format, your audience can — and likely will — become fatigued or bored. Gradually, the tactics that once served you well will lose their efficacy.
This goes back to my earlier points. Data, AI, and analytics are great for providing suggestions, but they’re no substitute for human judgment or creativity. Even if you don’t automate your creative process, you shouldn’t outsource your thinking to a computer.
Automated Tools are for Busywork
I’ve talked extensively about the places where you shouldn’t use automation — or, at the very least, use them cautiously and sparingly. And from that, you might infer that I’m deeply skeptical about the role technology can play in the daily marketing grind. The truth is, I think there are plenty of opportunities where automated systems can shine.
The examples will vary from place to place. Many of the tasks once performed by a junior employee or an intern — particularly when it comes to project management or metrics-gathering — can be reliably offloaded to a machine. Automated systems are great at performing complex number-crunching tasks, like sentiment analysis and demographic targeting.
And while I’d be reluctant to let an AI write ad copy, I’d be happy to let it tackle tasks that are time-consuming, repetitive, and formulaic. Jobs that, although important, are often low-stakes and fall under the broad umbrella of “busywork.” That's why our team launched an AI campaign planner tool, amongst other features.
Again, these will vary between teams and organizations. Marketing leaders need to take a candid look at their own operations and look for ways technology can streamline the day-to-day. While this might meet internal opposition from those afraid of being made redundant by technology, the reality is that you’re simply freeing their time from mundane tasks so that they can focus their energies on the creative jobs that deliver real business value.
Marketing Teams Are Now Tech Teams
The last — and possibly most important — point is that adjusting to the pace of technological change in marketing requires a fundamental cultural shift. Marketers are no longer purely creatives, but also analysts and technologists. They need to understand the tools they use, learn how to use them effectively, and recognize their strengths and limitations.
That’s not to say that every marketer needs to be a programmer. But they need to embrace the fact that their roles are increasingly tech-centric. Just like a mechanic understands the workings of an internal combustion engine, they should learn how their organization’s AI and data tools operate.
This cultural shift brings several key benefits. First, there’s the day-to-day. On a basic level, people are happiest when using tools they understand and feel safe around.
The others are less immediately noticeable, but still important. This ongoing infusion of tech into the marketing world means that marketers will inevitably need to work closely with engineering and data science teams. For this collaboration to work, both sides need to understand the technology, its capabilities, and its limitations. They need a shared vernacular, and marketers need to be able to communicate their objectives to those beyond their world.
And there are psychological benefits to treating marketers like tech professionals. Engineering is fundamentally rooted in experimentation. A programmer will write a solution, run it, and measure whether it works. If it fails, it fails. Rather than treated as a catastrophe, failure is seen as an inevitable part of the job, and a milestone on the road to a working solution.
Marketers can learn a lot from that. Our reliance on data means that we’re often optimizing for success, rather than looking for new opportunities and flexing our creative muscles. And that has an inevitable opportunity cost.
We’re at a point where organizations are expected to be “data-driven.” Everyone aspires to be the team from Moneyball, and those still relying on instinct and experience alone are regarded as archaic and obsolete.
The problem is that “data-driven” is a shallow, unnuanced term. It doesn’t really cover the how, the why, or the extent. For marketers, there’s a real risk associated with letting machines do too much. Technology is no substitute for creativity or imagination, and it doesn’t understand the fundamentals of how audiences engage with the content they consume.
Moderation — and balance — is key.