top of page
Jahan Thakur

AI and Machine Learning in Advertising: the Human Factor



machine learning in marketing

*This is a continuation of our previous article, AI and Machine Learning in Advertising.


Machine learning has become a transformative force in many industries, including advertising. It’s undoubtedly capable of outperforming humans in certain tasks, especially those requiring data processing and predictive analytics. It can easily beat a novice at chess as early as 1956 and famously defeated Garry Kasparov, the reigning world chess champion, in 1997. However, as Frederick Vallaeys states in his excellent book Unlevel the Playing Field, “humans are better at context, nuance, and creativity.”


Machine Learning vs. Human Intelligence in Marketing


One of the strengths of machine learning is its ability to make sense of vast amounts of data and automate decisions based on patterns and trends. But it has its limits—especially when things go wrong or require a deeper understanding of human behaviour.


Let’s say the algorithm notices a decrease in conversion rate and reduces the bid accordingly. On paper, this seems like a sensible response. However, if the conversion rate dropped due to a technical issue (e.g., a landing page not working properly), reducing the bid would accomplish nothing to solve the problem. In contrast, an advertiser that noticed the problem during their periodic audit could ask the web development team to implement the fix accordingly. Without human oversight, identifying the root cause of the problem may take more time, which will disrupt the smart bidding process and could potentially lead to a long-term decline in sales.


If the system had understood the context behind the drop, it could have informed the advertiser of the specific issue and allowed for quicker human intervention and resolution. In today’s AI-supported marketing world, machine learning excels at identifying what is happening but still lacks the ability to explain why. This is where the Human Factor becomes essential.


What is the Human Factor?


To quote Frederick Vallaeys again: “Humans + Machines > Machines Alone.”

The author breaks down the roles of humans working with machines into three use cases: Teacher, Doctor, and Pilot.


1. Teachers: Guiding Machine Learning for Optimal Performance


  • Role: A teacher’s job is to set the machine learning algorithm up for success by structuring campaigns in a way that enables the technology to perform at its best.

  • Example: When building a lead generation campaign, human marketers must design the campaign architecture, define clear goals and ensure proper data integration. Without feeding CRM data back into platforms like Google or Facebook, the machine will be unable to understand what constitutes a qualified lead and therefore only optimise toward lead volume. By providing the correct data signals (such as which leads convert into actual sales), teachers ensure that the algorithm is optimising for the true business goal: revenue, not just lead generation.


2. Doctors: Diagnosing and Troubleshooting Issues


  • Role: When something goes wrong, the doctor’s job is to diagnose the issue, provide an explanation and help resolve it. They act as a troubleshooting guide for both the algorithm and the client.

  • Example: To return to our earlier example, if conversion rates suddenly drop due to a landing page error, it is the human doctor’s responsibility to identify the underlying cause. Machine learning can flag a drop in performance but cannot explain why it happened. A human marketer, however, can investigate whether the issue stems from technical problems, changes in user behaviour or external market conditions, and then recommend appropriate action.


3. Pilots: Monitoring Performance and Taking Control When Necessary


  • Role: Pilots are responsible for overseeing the performance of campaigns and stepping in to make critical adjustments when the machine fails to account for nuances in the market.

  • Example: If there’s a temporary issue on the website, such as a broken payment gateway, a human pilot can quickly pause ad spend or reallocate the budget elsewhere to avoid wasting money on traffic that won’t convert. Pilots also look for opportunities that algorithms might miss, such as increasing bids for ad placements that are under-performing due to low search rank or tweaking a creative that is under-delivering. While machine learning handles routine optimisations, it is the pilot’s job to intervene when deeper insights or immediate action are required.


4. Strategists: Thinking Beyond Tactics and Towards the Future


While the roles of teacher, doctor, and pilot address the day-to-day tactical aspects of digital marketing, the strategist looks at the bigger picture.


  • Role: A strategist is concerned with long-term growth and the alignment of marketing efforts with business goals. They don’t just respond to immediate issues but plan campaigns to maximise ROI over time.

  • Example: Strategists will often use a combination of tools such as scripts, third-party platforms and competitive analysis to identify opportunities for growth. They’re responsible for understanding the full customer journey, developing messaging that resonates with target audiences and uncovering missed opportunities that machine learning alone wouldn’t detect. For example, a strategist might identify a new trend in customer behaviour that requires a pivot in the campaign’s direction or use data analysis to predict where future opportunities lie.


A strategist also integrates creativity into the marketing mix, whether it is through compelling ad copy, fresh imagery or interactive content. Machine learning can suggest optimisations based on historical data, but it cannot create the kind of innovative, emotionally resonant campaigns that build strong, lasting connections with customers.


Enhancing Machine Learning with Human Expertise


Human marketers can ensure that machine learning algorithms are set up correctly, troubleshoot problems when they arise and step in when a deeper understanding of human behaviour or market dynamics is required. The best results come from a hybrid approach—where machines handle the heavy lifting of data analysis and optimisations while humans bring the creativity, intuition and strategic foresight that guide long-term success. The companies that do this effectively will not only keep up with changes in the market but will also drive innovation and outpace their competitors.


If you have any questions, need advice, or want to take your digital marketing to the next level, we’re here to help. Our team offers a free audit to assess your current performance and show you how we can improve your results. Reach out today and let’s discuss how we can work together to maximise your ROI!

Comments


bottom of page