Expert Mode: Escaping the Generative AI Sea of Sameness with Toby Coulthard, Chief Product and Growth Officer at Jacquard
This article was based on the interview with Toby Coulthard, Chief Product & Growth Officer at Jacquard by Greg Kihlström, AI and MarTech keynote speaker for The Agile Brand with Greg Kihlström podcast. Listen to the original episode here:
The generative AI arms race is in full swing. Across the enterprise, marketing teams are rushing to deploy these powerful new tools, chasing the promise of unprecedented efficiency, scale, and personalization. The initial business cases wrote themselves: generate a hundred subject lines in a minute, draft campaign copy in seconds, and scale content creation beyond what any human team could ever hope to achieve. Yet, as the initial dust settles, a certain unease is creeping into the C-suite. The promised land of hyper-personalized, unique content is starting to look suspiciously uniform. A strange, uncanny valley of sameness is emerging, where brand voices, once painstakingly crafted, are beginning to sound eerily similar.
This is the central paradox facing marketing leaders today. The very technology meant to help us connect with individuals on a massive scale is, for many, simply amplifying mediocrity. By pulling from the same vast, public data pools and using the same foundational models, we risk creating an echo chamber where differentiation is lost and brand soul is eroded. The path forward isn’t to retreat from AI, but to evolve our use of it. It requires a fundamental shift from a purely generative, reactive model to one that is predictive, strategic, and deeply integrated with the unique DNA of your brand. It’s about moving beyond simply creating *more* content to creating the *right* content—the content that is not only on-brand but is also verifiably predicted to perform.
The Problem with Canned Creativity
The root of the problem is simpler than many realize. While there are numerous large language models (LLMs) available, the vast majority of marketers are defaulting to the most accessible one. This creates a feedback loop of homogeneity. As Toby Coulthard points out, the issue is not just that the models are converging, but that our usage of them has converged first.
“The reality is 85% of marketers are just using Chat GPT to… generate content. And so it’s we’re not even talking about using some of these other models… We can all kind of feel AI-generated content when you see it. If you go on LinkedIn, there’s a certain thing about it, there’s a tone of voice, there’s a cadence… you can just tell that it’s AI written. And there’s this convergence of different people sounding the same.”
This isn’t just an aesthetic annoyance; it’s a significant business risk. For decades, enterprise brands have invested millions in defining and defending a unique brand voice. These carefully constructed guidelines are a core asset, the very essence of how a company communicates its values and differentiates itself in a crowded market. Yet, in the rush for efficiency, many teams are inadvertently bypassing this asset. They are trading distinctiveness for speed, a bargain that may boost short-term productivity metrics but will ultimately dilute brand equity. The danger, as Coulthard notes, isn’t just that your marketing becomes less effective; it’s that it risks becoming completely ineffective, blending into a background hum of generic, AI-generated “slop.”
Moving from Testing the Bad to Predicting the Good
The “test and learn” mantra is deeply ingrained in the culture of digital marketing. It has served us well, allowing for data-driven iteration and optimization. However, when paired with generic AI content generation, this trusted methodology reveals a critical flaw. If you ask an LLM for ten email subject lines, it will give you ten grammatically correct but likely uninspired options. Your A/B testing will then diligently identify the best-performing of those ten options. The problem is, you may only be discovering the best of a bad bunch.
“The test and learn thing, that’s still very much should be part of a marketer’s toolkit. But what you don’t want to do is just test a bunch of bad variants of language, right? Like it’s you’re just finding the best of a bad bunch. I think at the end of the day, you want to be able to… put your best foot forward. You want to test a range of diverse but also performant variants.”
This is where a strategic shift to a predictive model becomes essential. Instead of generating content and then testing it, a more advanced approach uses a different layer of AI to predict performance before a single piece of copy is deployed. This isn’t the same kind of prediction an LLM performs when it guesses the next word in a sentence. This is prediction based on a model trained specifically on a brand’s own historical performance data—billions of data points from opens, clicks, conversions, and downstream metrics like lifetime value. This predictive layer acts as an intelligent filter, sorting the wheat from the chaff and ensuring that the content that makes it to the “test and learn” phase is already primed for success. It’s a move from hoping for a winner to engineering one.
The Elevated Role of the Human Expert
A common narrative surrounding AI is one of replacement, particularly for creative roles like copywriting. However, a more sophisticated view sees AI not as a replacement for human expertise, but as a powerful force multiplier that elevates it. The copywriter’s role doesn’t disappear; it evolves from being the “machine” that types the words to being the “strategist” who directs the machine. Their deep understanding of brand voice, narrative, and customer psychology becomes more crucial than ever.
“I don’t think any copywriter needs to be an expert in typing, right? For example… they’re still a very much necessary part of the process, but they’re the ones using the machine rather than being the machine, so to speak.”
In this new paradigm, the copywriter becomes the conductor of an AI-powered orchestra. They set the strategic direction, define the brand guardrails, and fine-tune the AI’s output. The AI, in turn, handles the tasks that are impossible at human scale—generating thousands of hyper-segmented, personalized variants for countless customer journeys, channels, and touchpoints. This frees the human expert from the menial aspects of content production and allows them to focus on higher-level strategy, creative oversight, and ensuring the brand’s soul remains intact. The result is a symbiotic relationship where human creativity guides machine-level scale, achieving a level of personalized, on-brand communication that was previously unimaginable.
Navigating Nuance: From Compliance to Culture
For enterprise marketing leaders, scale introduces complexity far beyond just creating more content. It involves navigating a minefield of regulatory compliance, cultural sensitivities, and geographical nuances. A generic LLM, trained on the open internet, is simply not equipped to handle this level of specificity. It can hallucinate, make unsubstantiated claims, or produce content that, while technically correct, is culturally tone-deaf. This is where the combination of generative AI with deterministic, rule-based systems and predictive models becomes critical.
“In the UK, people do not like content that is highly urgent. They want calm, helpful language. They don’t want to have something that says sale ends in 24 hours, buy right now. But in America, that really resonates. Right. And so understanding that and training a model that understands the nuance… can be the difference between really being successful in a brand and not.”
Whether it’s avoiding specific health claims in pharmaceutical marketing or understanding that urgency-driven language performs differently in London than it does in New York, a truly intelligent system must be trained on these subtleties. By layering “neuro-symbolic guardrails” over a generative engine, brands can ensure that all content is not only performant and on-brand but also safe and compliant. This framework reinforces the idea that not sending the wrong message is often more important than sending the right one. It’s a level of sophisticated risk management that simply can’t be achieved by prompting a public-facing chatbot.
The initial wave of generative AI adoption has served as a valuable, if sometimes humbling, learning experience. It has exposed the limitations of a purely efficiency-driven approach and highlighted the enduring importance of brand differentiation. The leaders who will win in the next phase are those who recognize that generative AI is not a standalone solution, but a powerful component within a more comprehensive marketing intelligence engine. It is the combination of generative creativity, predictive analytics trained on your own proprietary data, and the irreplaceable strategic oversight of human experts that will unlock true competitive advantage.
This evolution is about more than just better technology; it’s a shift in mindset. It’s about moving from asking, “How can AI help us do things faster?” to “How can AI help us understand our customers better and communicate more effectively?” As we head into a future that will undoubtedly be saturated with “AI slop,” the ability to cut through the noise with a message that is authentic, relevant, and verifiably human will be the ultimate differentiator. The greatest advantage will belong not to those who adopted AI first, but to those who apply it most intelligently.
