In 2003, Seth Godin introduced the idea of the “purple cow” in his iconic book “Purple Cow: Transform Your Business by Being Remarkable”. He used the idea to describe how consumers are constantly bombarded with an overwhelmingly large number of choices and ads, making it difficult for anything to stand out.

Godin compared this to seeing a field full of ordinary cows. At first, the sight might be interesting, but after a while, the cows become indistinguishable and unremarkable. However, if you spotted a purple cow among them, you would immediately notice it. Its completely unusual appearance would make it memorable and remarkable.

AI slop is the modern version of the problem of indistinguishable cows, except that it is far more severe. One can keep churning out software, digital products, or content without the hurdles of capability, bandwidth, or time. You can have a one-person-led factory assembly line producing content or software, beating the output and scale of large companies from just a decade ago.

And this extreme volume of AI slop can easily bypass differentiation or plagiarism checks, as AI can be trained to modify core original IP or ideas into a million copies that all look different from the original or from each other. And yet, customers instinctively recognise AI-generated content or products when they see them, and they end up tuning them out. So how do you break through this blindness and capture customer attention?

Introducing the Green Unicorn

In that very Seth Godin-esque proverbial field that is now full of purple cows, imagine seeing a green unicorn. It would absolutely catch your eye and be remarkable enough to be talked about with others. For eyes trained to see the possible, the sheer impossibility of a unicorn is remarkable enough. But making it green adds another layer of distinction. It avoids even the very familiar pink-and-purple imagery we have come to associate with unicorns from fairy tales and children’s books.

If a purple cow is unusual, a green unicorn is extraordinary.

If a purple cow can be made possible and replicated, a green unicorn continues to be fantastically impossible to create or replicate.

How to Be a Green Unicorn

A purple cow is a unique version of a common sight in real life(the cow). A green unicorn is a unique version of an imaginary sight. The hindrance to having a purple cow is one’s ability to imagine one. Once it has been imagined, however, it is remarkably easy to have many purple cows (Purple-dyed cows are a thing).

To be a green unicorn, however, you have to break out of the frame of reality. Here, reality refers to what is possible within the universe in which the green unicorn is set. The hindrance to having a green unicorn isn’t just one’s ability to imagine one. Making it happen is a much bigger hindrance and the person who makes it happen will continue to have many more of them, whether they are green or any other color.

Green Unicorns in Real Life

To become a green unicorn in your domain, you have to step outside the established frame of what is considered “reality” in that field. AI systems identify and learn patterns from the data they are trained on. They are effectively modelling a statistical representation of the realities captured in that data. Anything that falls outside that distribution is close to impossible for AI to reproduce reliably. And what has no established pattern to draw from is unlikely to become another predictable, mass-produced variation.

For instance, if you are a maker of digital products, you have to come up with a category that hasn’t been named or identified yet. Once identified, that category shouldn’t be easily replicated without unique access to the data needed to create it.

Let’s say you are a developer who builds software products. Your green unicorn needs to break out of known frameworks like “Customer Relationship Management” or “Marketing Automation” and look for types of software that require special access to data and are too bespoke to be easily replicated.

- Custom project status report that pulls data from Jira, Slack, and project-management systems to produce the team’s bespoke project status report, highlighting the current status of each project. Except that it also incorporates the team’s unwritten and confidential definitions of status values such as “on track,” “at risk,” and “critical”(definitions that rely on subjective interpretations of industry regulations). You know these definitions because you work on the project.

- Custom retail inventory report that includes a custom “stock on hand” field, calculated using a specific combination of current stock, sales velocity, upcoming promotions, and supplier lead times. This is a requirement they share with you because you previously built a tool for them that tracks upcoming promotions.

- Product team’s “Friday product pulse” that combines customer feedback, support tickets and product analytics, then classifies them according to the team’s own categorization rather than a standard product-management framework. You get to build the tool because the team approached you with the requirement.

If you are a content creator, you can no longer let Google or ChatGPT be your primary discovery source. This is especially true for content creators whose content is more informational or idea-based, as opposed to entertainment-focused. Your ideas, expressed through your content, will simply be ingested, aggregated and mixed with content from other sources when AI answers the questions users ask it.

What you need to be doing instead is releasing video and audio content based on direct, first-hand experiences. These can be podcast interviews, behind-the-scenes experiments, original research, customer conversations, case studies, personal observations, or documentation of something you are actually doing.

The more your content captures experiences, insights and evidence that cannot be easily replicated from existing information, the more you can present a version of reality that AI cannot directly access. Over time, users may start coming directly to you as the primary source, whether to stay updated on what’s happening in your domain, hear your perspective on it, or simply find out what your body of work says about a particular subject.

For instance, product management podcaster Lenny Rachitsky has his own AI chatbot, LennyBot, which answers product-management-related questions based on content from his podcasts, blog, and newsletters. Responses from LennyBot to product-management-related prompts are much better than responses from general-purpose AI tools such as ChatGPT because of the highly focused and experience-rich nature of Lenny’s podcasts and newsletters.

Break the Frame to Become a Green Unicorn

In 1975, Queen released a song that challenged almost every rule of popular music at the time. It had no conventional chorus and shifted between ballad, opera and hard rock. It also ran for nearly six minutes, far removed from the usual radio-friendly three-minute format. EMI was initially hesitant to release such an unconventional song as a single, fearing that radio DJs wouldn’t play it. Yet Freddie Mercury and the rest of Queen resisted the pressure to make it shorter, simpler or more familiar because they wanted to create something that sounded like nothing else on the radio. Decades later, that courage and conviction to be different is precisely what made the song so unforgettable.

That song was *Bohemian Rhapsody*, which went on to achieve more than 10 million sales and stream equivalents in the U.S. alone and has been recognised as one of the greatest rock songs of all time.

AI couldn’t have come up with a song like “Bohemian Rhapsody”, nor can it create anything remotely as good or as popular as the original.

Green unicorns will always be the answer to standing out from and rising above mass-produced clutter.