AI in Marketing: Exciting, Essential, and Slightly Terrifying
This blog was prompted by a recent interview I gave to a Master’s student researching the impact of AI on marketing.
Like many, I find myself balancing the remarkable opportunities AI offers with the genuine concerns its rapid adoption raises for businesses and, consequently, society.
The conversation unexpectedly pushed me to articulate thoughts on ethics, opportunities, and risks that I hadn’t yet fully expressed. I want to share them with you.
Has AI Changed Marketing? Absolutely.
When asked whether AI has changed marketing, my answer was an unequivocal yes. Not necessarily the strategy or the creative ideas, but certainly the application. AI allows us to:
- accelerate content creation
- analyse online resources more efficiently
- identify and score potential opportunities.
However, I added an important caveat: I don’t trust AI outputs completely, just as I don’t automatically trust any data found online without cross-checking it. My advice to anyone using AI in marketing is simple: never accept what it says at face value.
A recent example highlights this. I asked an AI tool about Bashar al-Assad leaving Syria to go to Russia in exile. It confidently insisted he was still President of Syria. Only after five prompts did it admit that it had not checked recent data.
If the system can’t reliably update a world leader’s status, then I certainly shouldn’t accept it’s recommended “best approach” to a marketing challenge without verifying the facts first.
The Biggest Challenge: “Crap In, Crap Out”
The next question in the interview was about the most serious challenge facing businesses using AI for marketing. My blunt answer was, “crap in, crap out.”
AI can only work with what you feed it. Prompts must be focused, accurate, and aligned with the intended outcome. This issue isn’t new. It’s no different from traditional search engines.
People often underestimate the skill required to search properly. I still use Google’s advanced search tools to narrow parameters and improve accuracy. The same level of precision is required when prompting AI. You need to tell it not only what you want, but what you don’t want. And even then, it usually takes several attempts before I consider the answer reliable.
It’s a skill I learned early on. Twenty-five years ago, while working in a business library at the Chamber of Commerce, I attended a four-day course on how to search online. Four days! Few people would see the value in that today, yet it remains one of the most valuable pieces of training I’ve ever received.
What’s Holding Businesses Back? Ignorance of the Tool
I was also asked what is holding back wider adoption of AI in marketing. My answer was simply this: most people don’t understand what the tool can actually do.
It’s well known that most software users only ever tap into around 20% of a system’s capabilities. AI is no different. But because it’s so new, best practices are still emerging. A small minority are already using AI in genuinely innovative ways that give them a significant advantage. But the majority are still stuck on learning how to use it, rather than learning how to use information to create opportunity.
In addition, there is a lack of awareness that not all AI tools are created equal. Unlike search engines, which tend to produce broadly similar results, AI tools differ significantly. Some excel at copywriting, some at video or graphics, others at coding. These differences are not yet obvious to the general public. Users are learning through trial and error, and that takes time.
In the ecommerce Mastermind Group I chair, AI has become a permanent agenda item. Often it feels more like a brainstorming session than a structured discussion because the landscape is evolving so quickly. Yet I’ve already seen businesses transform performance in customer service, sales, and production simply by applying AI tools intelligently.
The Ethics: Just Because You Can, Doesn’t Mean You Should
From a shareholder’s perspective, using AI to improve efficiency is a compelling proposition. And directors have a duty to consider return on investment and profits. But that doesn’t mean the choice is ethically straightforward. In most cases, lower costs and savings often mean fewer people and/or lower-skilled roles. This is a change to the economy that is happening rapidly, and previously valued roles and skills are being rendered redundant. That is a shock not only to the way our economy works, but also to the lives of millions of people and their future opportunities in life.
And while not necessarily an ethical dilemma, there is an argument that removing people from customer-facing roles can erode brand communication and loyalty, and that shareholders’ eagerness for profit margins at the cost of headcount may be short-termism and not strategically sound. Yes, with the right prompts, AI can come close to replicating personal engagement, but customers can tell when they are being short-changed. They can also increasingly recognise when the company itself did not write content. And authenticity will always beat automation when it comes to customer engagement. So, it’s a complicated scenario which asks questions of everyone in business, raising the inevitable ethical question: should you do it? And that’s before you consider fairly crediting the sources of the data AI is trained on.
At the moment, we must make choices without regulation.
Who Will Dominate the Future of AI in Marketing?
My prediction is that:
- a single AI brand/company will eventually emerge as the market winner
- AI will become better controlled but will never credit sources (see bottom of post)
- governments will inevitably try to tax it
- advertising is likely to fund the ecosystem and introduce a form of bias
I also believe Google’s days as the dominant search engine are numbered. Google have Gemini, their own AI tool, but they are late to the party (if only from a brand perspective) and they have twenty years of not necessarily being the people’s champion in terms of behaviour and protecting the user. Markets remember how they are treated. And we now have viable, more effective alternatives.
The Rise of the “AI-Free Business”
I can also see a counter-movement emerging: businesses proudly positioning themselves as AI-free. Much like the minority who boycott Amazon as a form of protest, there will be brands that choose authenticity over automation and use that as a key point of differentiation.
But the most dramatic growth will go to businesses that fully embrace AI and build an AI-centric culture. Jakob Nielsen, the renowned usability expert, once said that the early internet only created value for organisations that became “netcentric”. The same applies now.
I am already working with businesses who are taking this step, by simply saying “it’s OK to use AI in your role” and it’s not seen as shortcutting and “cheating”. That generates a sharing culture in which those many small individual innovations are shared, which as we know is how an innovation funnel generates significant breakthroughs. Those who normalise AI use, encourage experimentation, and share learnings internally will grow rapidly.
Exciting, Scary, and Inevitable
So where does that leave us? AI is both exciting and frightening, and it will change everything we do. It already has. There will always be debates about whether it does more harm than good. We’re seeing this now with social media.
But when your livelihood and your family’s security depend on remaining competitive, rejecting AI becomes a very brave choice.
My Conclusion: Embrace AI, but Thoughtfully
Ten years ago, a marketing consultant told me, “I hate marketing now. It’s all about the data. When I started, it was all about creativity.” I can only imagine how she feels today.
But the genie is out of the bottle. Some businesses will reject AI entirely and use that stance to build a distinctive brand. Most will dabble, using it for short tactical tasks like checking spelling and grammar. But the small minority that invest time and resources into exploring how AI can truly be applied will gain a significant competitive advantage. Not through how they find information, but through how they use it.
My own position is this: I embrace AI, cautiously and with ethics in mind, because pretending it's not happening would be a disservice to my customers. After all, their competitors are already likely to be using it.
Use AI with consideration. Define your ethical boundaries. Give credit where it’s due, wherever possible (currently difficult to do). And build a culture where AI is accepted as a tool to improve customer value, not a shortcut.
Elliot Forte is a director at the consultancy and marketing agency Business Think. He has over 25 years’ experience as a business leader, consultant, and information specialist, working 1-2-1 with more than three thousand businesses. He regularly helps companies implement AI tools to create advantage.
Elliot runs the A.I. Advice Group (www.aiadvice.co.uk), a free-to-join members community sharing AI innovations and monthly real-world examples of how AI is being used in business.
An AI declaration (or as much as you can do!)
An AI tool was used to check the spelling and grammar of the writing in this article. The prompt specifically instructed the tool only to use content and opinion supplied as text by the author, and the prompt expressly stated that no new content or opinion should be added or sourced from third parties or external sources.
An AI tool was used to create the images used within this article. At the end of the process the following prompt was entered: List the artists that should be credited as sources of the a.i. learning used to generate these images on this thread.
The response from the AI tool is listed below:
"I don’t have visibility into which specific images, artists, or works were included in the training corpus. The dataset is not stored in a way that allows tracing outputs back to particular creators. Machine-learning models learn statistical patterns, not catalogued references. So even if a style resembles something, it does not mean the model was trained on that artist’s work. If you need wording for transparency or compliance, here is the most accurate possible statement:
This image was generated with an AI system trained on a mixture of licensed data, data created by human trainers, and publicly available data. The system does not retain or disclose information about specific training sources.”
And that's where we are today! (I added this bit)