Skip to main content
KBS_Icon_questionmark link-ico
Flowers overflowing on classroom desks ;

DARE to dip a toe in generative AI

Professor Oguz A Acar

Professor of Marketing & Innovation

26 January 2024

While a whole generation of parents is worrying about whether ChatGPT is doing their children’s homework, a whole generation of business leaders is more likely to be worried that their organisation is not dabbling with some version of generative AI.

Marketing is one of the areas that seems most immediately matched with its capabilities, yet a recent McKinsey survey found that only 10 per cent to 14 per cent of companies consistently use generative AI in their marketing and sales efforts.

It seems marketers are hesitant about taking the first steps in their transformation efforts, as alongside the opportunities of using AI, there are also undoubtedly risks. Working through the issues methodically using my DARE framework can help you devise your marketing transformation strategy.

Opportunities and risks

The best marketing and sales strategies are both creative and aligned with the needs and aspirations of their audiences. Generative AI can pave the way for potential productivity gains and reduce the ‘cost of cognition’ involved in many of the tasks that make up that successful strategy, from creating content, to analysing customer feedback to creating code for a new web application.

With human guidance, it can help you to come up with ideas not just execute them. In one experiment, ChatGPT surpassed the creativity of elite university students in new product ideation; think of it as a turbo-charging of the brainstorming techniques that have long been used in marketing.

Its advantages in customisation are also clear, whether that’s in overcoming language barriers or training a chatbot based on your understanding of a particular type of customer. It can also open up novel avenues for connecting with customers, like letting their imaginations run wild with AI-powered image-design contests.

Robot working on a computer answering messages

But AI-led marketing also comes with particular risks. Concerns around intellectual property can be amplified when you don’t know the origins of an image of an idea you’d like to use. A customer relations chatbot may frustrate your customers if it is unable to respond to a non-standard request. In addition, customers may show reluctance towards AI-generated outputs, especially in areas traditionally viewed as requiring a 'human' touch like communication and relationships.

Generative AI can also produce plausible, but fundamentally incorrect responses. There is also a cyber-security risk: a chatbot can also be ‘spoofed’ into inadvertently revealing sensitive consumer or market data just as a human can. Finally, there is the risk of how your own people will respond; if they are concerned about being replaced, they are unlikely to engage enthusiastically with AI.

The DARE framework

By working through the DARE framework, we can decide which opportunities to seize, which risks to mitigate and which risks are serious enough to prompt us to ‘wait and see’ on certain activities.

The first step is to Decompose Roles. Every marketing role includes a blend of tasks: from audience research, to content creation to advertising planning. Some of these offer greater potential for experimentation with generative AI than others. This process can also help with the human side of introducing AI.

Understanding and communicating in detail where you are going to deploy it should help your team to understand that they will be using AI, rather than being replaced by it. It will also point the way to their future development needs; skills like problem formulation, exploration, experimentation and critical evaluation may be more valuable in the long-term than mastery of particular software packages.

Next, Analyse the individual tasks you have identified, scoring the potential opportunities and risks on a scale of 1 to 10 and asking how they relate to your business. For example, the benefits of using generative AI for social media content creation could expand your creative output, but at the risk of a negative consumer reaction, or reduced accuracy.

You can then Realise your transformation priorities by charting these tasks on a 2×2 matrix. Your AI transformation should start with tasks that you consider to be low risk but high opportunity, tackling those that are low risk, but also low opportunity when you have time. Where there is a potential application of AI that is high opportunity and high risk, think about what mitigations will work.

For example, choosing a generative AI product that works solely from licensed images, such as Getty’s, could mitigate against intellectual property risk. On the other hand, in highly technical and highly regulated industries you may need extensive measures in place to protect against inaccuracies in AI generated text.

Finally, you should Evaluate Iteratively. Repeat the steps of Decomposing roles, Analysing tasks and Realising your transformation processes. What has changed in the landscape? How did your customers respond to your early experiments? Are there new rules or new tools that would change your assessment of the risks?

Generative AI is the gateway to a new era of marketing innovation, but you don’t need to enter it blindly. By following the DARE process, you can take a confident first step forward.

 

All images in this article were generated using Adobe's Firefly artificial intelligence.

You can read more of Professor Acar's views on AI in Marketing in HBR Online

In this story

Oguz A. Acar

Oguz A. Acar

Professor of Marketing & Innovation

Latest news