What AI can’t or shouldn’t replace humans for in the business of marketing

by Megan Arnold

In the application and integration of artificial intelligence into daily business processes, there’s understandable and legitimate angst over how AI will replace humans. With what we understand today about AI, I believe that these new capabilities will let us be more human, not less. I believe that AI will remove the kinds of tasks from our to-do lists that give us more time to be the creative, inspired, intellectual, metacognitive, and talented humans we all are. 

When AI can automate a number of our daily tasks as marketers- like writing swipe copy, first drafts, scheduling meetings, using templates, etc.- we are left in the position of focus on critical management functions. These management functions aren’t relegated to managers or executives, every marketer makes strategic decisions at every level. As we work towards integrating AI into our production processes, understanding which functions remain owned by humans will help us all establish the roles and responsibilities of our newest team member - artificial intelligence. 

Here are a few management functions I see as primarily human:

Corporate strategy

AI will inevitably have a large role to play in the development of corporate strategy given its ability to analyze massive amounts of data quickly. AI algorithms rely heavily on historical data to predict future outcomes, and only on the historical data it has access to. In the near term, AI will remain blind to competitor data or data proprietary to entities outside the firm.

Prioritization

One fundamental responsibility of managers in executing corporate strategy is deciding what not to do. Managers are often faced with many options, many potential paths forward, many good campaign ideas to invest in, but the purpose of strategy is to maximize the results produced while minimizing the resources required. This involves saying no more than it involves saying yes. In an AI-driven world where so much more marketing work can be produced than ever before, the temptation to say yes more often will be strong. It is essential that managers remain committed to using AI production processes to drive efficiency in achieving strategic objectives, not in losing focus.

Decision making

While AI tools can provide recommendations on what decisions to make (like project prioritization) based on the resources required and the results expected, there are a number of factors involved in decision-making that can’t be reflected in data. These factors could include information humans would know that AI doesn’t, like a new product in development, anticipated changes in the market, unprecedented economic environments (like a global pandemic), or personality and power dynamics in a business.

Humans should continue to make major decisions in the marketing production process, the most important of which may be what gets published, or which campaigns go live. Marketing artifacts represents a company’s external image and reputation at scale, and artificial intelligence should not automatically publish those artifacts without human approval and oversight.

Quality assurance

AI is not perfect and cannot produce perfect results. Perfect is subjective, as is the definition of quality. In marketing terms, quality can mean anything from an asset’s adherence to brand style guidelines to technical accuracy or to just being fun to read. Generative AI tools can take into account many of the technical aspects of writing and content, even brand tone of voice, but it may not catch bias baked into its work. Humans must remain gatekeepers to AI-produced work to prevent unconscious bias, inaccuracies, or that the finished product matches the brief and follows through on its strategic intent. 

Innovation and original works

Marketing is inherently creative. It’s the practice of building meaningful connections between an individual and a brand at scale. When we operate at scale, we can identify trends and derive insights applicable to most of the people engaged with our campaigns, but each of those people remain unique, complex individuals looking for solutions to their specific problems. On the other side, each corporation sees itself occupying a unique position in the market relative to its competitors. It invests in producing new solutions through innovation, in bringing products to market that customers haven’t seen before. 

Today, generative AI relies on data and information that already exist and can be processed by machines. The complexity of the human individual can’t be entirely captured by AI (fears, regrets, hopes, insecurities, to name a few), and the ability to imagine what has never existed before remains distinctly human.  

As long as we see the production of new ideas as informed by AI but owned by humans, our markets will remain a step ahead of AI.  For marketers, this means that human talent should focus on the innovative work of producing the knowledge base the AI tool relies on (messaging frameworks, technical documentation, brand guidelines, etc.) and experimentation with new tactics to continue informing AI on best practices.

A culture of AI literacy, healthy skepticism, and a commitment to ethics

In an AI-driven production environment, marketers should be well-educated on what’s behind the algorithms - the knowledge base the tool uses, the rules it follows, and potential biases inherent in its outputs. When reviewing work produced by AI, marketers need to continually question if the asset reflects strategic intent, is free of unconscious bias, and represents the company appropriately. 

Ethical considerations have a necessary role to play in the development and adoption of AI tools. In marketing, some of those considerations include data privacy, systemic discrimination and unconscious bias, accuracy and veracity of information, corporate espionage (competitive intelligence), impersonation (deep fakes), and corporate social responsibility. 

As the application of AI evolves and becomes more understood, marketers can see AI as a powerful engine of information and analysis that enables us to be better managers. We maintain authority over AI as we leave critical management tasks to humans, from decision-making, to ethics, to innovation and new ideas.