I had an epiphany while trying to read a blog post the other day. Not just any brand’s blog, but a Fortune 500 company. A well respected company that had set the bar for “best-in-class” branding for dozens of years.
It started like this. At first I had a hard time focusing on the topic – I found myself re-reading paragraphs because they just weren’t sinking in. Next I skipped ahead, because somehow I knew what was coming next. I’ve read this blog before. A thousand times before. Just like Neo in the Matrix could see all of the green ones and zeros streaming down in his reality – I could see all of the data-driven decisions that crafted every phrase; I recognized the algorithms that chose every jargony word; and all of the industry best practices perfectly calibrated as they stared back at me. I was looking at “out-of-the-box” generative AI.
We’re all searching for ways to make our brands stand out, to be unique, to capture the fleeting attention of consumers. But the more we turn to AI for help, the more our efforts start to blend into homogenous digital vanilla.
So, the challenge for marketing teams is clear: how do you create something truly original when everyone has access to the same AI tools and techniques? How do you inject genuine personality into an algorithm-driven world? How do you make your brand’s AI stand out from the crowd when it’s built on the same foundation as everyone else’s?
Marketing is obsessed with data. Every click, every scroll, every purchase is meticulously tracked and analyzed. And while this data-driven approach has undeniably yielded valuable insights, it’s also led to a curious side effect: a creeping sense of conformity across brands.
When everyone’s basing their strategy on the same metrics, the same analytics dashboards, the same “best practices,” and now using the same AI, it’s no wonder our marketing efforts start to resemble one another. We’re all optimizing for the same outcomes, speaking the same language, and ultimately, producing the same predictable content. It’s the “Cult of Best Practices” taken to its logical, yet creatively stifling, extreme.
And then there’s the seductive efficiency of AI. These intelligent algorithms streamline our workflows, and make those data-driven decisions at superhuman speed. But when we become overly reliant on AI, we risk outsourcing our individuality to the machines as well.
When used in marketing, generative AI is like a magnet that pulls everything to the middle. This works well if you’re not very talented, and don’t want to put in a lot of effort – your work gets pulled up. But if you are talented, and use AI for efficiency – your work gets pulled down.
It’s not just writing, it’s strategy. The AI becomes the default, the path of least resistance, and the result is a sea of virtual brands all sounding eerily similar, their unique identities dissolving into a generic AI hum.
The next wave of AI solutions are poised to insert the same AI into every product and process with the assumption that it’s good enough as-is. Or at least one day soon it will be better. But do you really want to rely on a handful of engineers to decide what’s good for you?
Your brand is unique. Your AI should be as unique as you are.
How do you truly own a piece of AI, especially when it's something as ubiquitous as a public LLM? It starts by viewing the AI not as a tool – but as a collaborative partner. And like any good partnership, it requires a bit of give-and-take.
Here’s a good place to start. If your team directly prompts LLMs as part of your workflow, they may also be using popular prompting techniques designed to yield the same results. Don’t forget every time you ask for something compelling, resonating, or using best practices, you get exactly what everyone else gets. Simply stop using these common terms.
“In today’s ever-evolving business landscape…” has become the corporate equivalent of “Once upon a time.” First it's noticeable in the writing, then you see it in the strategy.
Flip the conventional wisdom on its ear. You don’t need to be a prompt engineer to get the most out of generative AI. You can invent your own terminology. The more abstract and unusual it is, the better. Here’s a hint: AI loves this. You’re teaching it something it doesn’t already know and it will pay close attention to the new rules. Prompt it to work with you as a partner to shape your own unique framework.
Reskill your team to become AI practitioners, willing to experiment. Through a continuous cycle of prompting, feedback, and refinement, your marketing team can imbue the AI with your brand’s distinct character, its quirks, its imperfections.
Consider building a curated “prompting lexicon” – a centralized repository of unique, brand-specific prompts that become the very DNA of your AI. Each successful prompt serves as a building block in your AI brand’s evolving personality.
I think the next logical step is building an AI agent or persona. At this point, I don’t have suggestions for what a marketing team can do internally without investing extensively in an AI-first knowledge management and automation system. Here’s a hint: this part is really hard. It's been the bulk of our work at Innovation Algebra for the past year.
Our focus is on creating AI personas that can conduct foundational research and strategic formulation, the kind of deep thinking that you traditionally get from expensive consultations. These bespoke AI personas can become invaluable partners for marketing teams, offering insights and recommendations that are not just data-driven, but also deeply aligned with the brand’s unique personality and goals. Each with their own distinct personality, knowledge base, skillset, and role to play in the business world.
It’s one thing to train an LLM to write a catchy tagline or generate a product description. It’s a whole new level to build a realistic customer persona, model a real business scenario, and have the persona suggest strategies and create materials that support optimal outcomes. This is the unique collaboration that we envision at Innovation Algebra. Moving away from one size fits all AI taking you a step closer to owning your own AI.
The next time you read an article or engage with a brand, look for the signs. Chances are you’ve read it before… Except this blog, which, of course, is totally unique.
Editor's note. As a reversal of roles, this blog was planned and outlined by our AI team and written by me. The personal stories and observations are mine.
