Why AI’s Jump to Conclusions Might Be Perfect for Your B2B Brand

Do you jump to conclusions?

It’s inevitable for humans.

Wait, did I just jump to a conclusion?

Psychology says cognitive biases motivate people to jump to conclusions. For example, association bias involves seeing connections in information where none exist. You reach an unwarranted conclusion based on a minimal set of data.

When I argue with my wife, association bias is the No. 1 reason why.

But can jumping to conclusions lead to good things?

B2B marketers must jump to conclusions

One of the most difficult – and yet ironically beneficial – aspects of B2B marketing is its focus on a niche audience. I once asked a marketing executive at an enterprise engineering company about his total addressable market (TAM). He grabbed a paper Rolodex and replied, “It’s the 200 or so companies in here.”

Statistical relevance is a hurdle for B2B marketers to ascertain what content resonates most with audiences, generates the most leads, and differentiates the brand. It’s not uncommon for even big B2B marketing teams to measure monthly web traffic in the thousands, leads in the hundreds, and monthly opportunities in the teens.

In the early 2000s, I was the chief marketing officer at a web content management software company. Our monthly goal could be creating or nurturing as few as 30 leads. The company would close an average of five to 10 new customers a month.

Understanding which ads, platforms, events, and thought leadership topics resonated the best hinged on a small number of people. We looked at the minimum data and estimated what worked. We had to jump to conclusions.

Now, some B2B companies jump to the right conclusion – the perfect thought leadership message or brand differentiation. The flywheel starts because differentiation happens quickly. By finding the groove of a disproportionate share of voice, marketing and sales become easier.

The perfect example of jumping to the right conclusion is the concept of inbound marketing.

Inbound marketing: A great jumped conclusion

In the early 2000s, an interesting trend in digital marketing appeared called “article marketing.” Brands could create interesting, thought-provoking articles on the web that would help the companies be discovered through search engines. Sound familiar?

But none of the content management or marketing automation solutions latched onto that as a messaging strategy. (To be fair, it wasn’t as obvious as it is now.)

In 2006, Brian Halligan and Dharmesh Shah founded HubSpot as a way to grade your website, look at social media engagement, and create blog posts and landing pages for leads. They coined the term “inbound marketing.”

This Google Trends graph shows that searches for “inbound marketing” (red line) gained its groove around 2008. It overtook searches for “article marketing” (blue line) by 2013. HubSpot had made the “inbound marketing” messaging a standard.

Brian didn’t have data on which to base that messaging strategy. If he had used the available data, he might have centered on “article marketing” as the term. But he looked at Dharmesh’s success through blogging and connecting through content on social media and believed that represented a new way of buying. Brian liked the concept of calling it “inbound,” as he shares in this 2019 interview.

Just about every B2B company I’ve worked with tries to find a flywheel like HubSpot did. But the challenge of a limited data set remains. Is it any wonder B2B companies have a seemingly constant and perennial “messaging strategy” evolution?

AI’s jump to conclusion may provide a B2B opportunity

I see an emergent challenge and perhaps a unique opportunity in generative AI and B2B marketing and content.

Generative AI tends to confidently “make up” answers. These “hallucinations” occur because the LLMs (large language models) acting as information sources are limited to what’s generally available on the internet. For niche B2B content, these sources may be few. So, when it comes to B2B content, generative AI often jumps to false conclusions.

A few years ago, I worked closely with a B2B client in the business of containerized data centers. Its exciting technology configures portable data centers in shipping containers, as you see behind a semi-truck cab. They can be stacked like LEGOs to scale quickly. The perfect use case is a natural disaster when emergency-response teams need an instant data center that can have electricity, store information, and allow for access to software, etc. These data centers can be set up in hours.

Now, this is a very niche B2B market.

Recently, I asked ChatGPT to describe the pros and cons of containerized data centers. The almost 400-word response listed three major pros and three major cons. Interestingly, the introduction outlined correctly that the future was bright for containerized data centers. Then, it wrote of general use cases, the ability of rapid deployment, and the reduction in time to reach difficult locales. It was correct though light on details.

But the other 85% of the response went off the rails. Every pro and con listed was confidently wrong.

This was the first pro: “Hybrid cloud environments: Containerization aligns well with the principles of hybrid cloud environments. Organizations can utilize containers to build and manage applications that seamlessly run across on-premises infrastructure, private clouds, and public cloud platforms. Container orchestration tools like Kubernetes facilitate this integration, enabling workload portability and efficient resource utilization.”

And this was the first con: “Complexity: Containerization introduces additional complexity compared to traditional virtualization. Managing containerized applications requires expertise in container orchestration platforms like Kubernetes and understanding the intricacies of container networking, storage, and security. It may require additional training and resources to effectively manage and troubleshoot containerized environments.”

Unless you’re in the industry, you may not see or understand how wrong that information is. The AI leaned on a different concept of “data containerization.” This data containerization is a software development approach in which the code is packaged up into a lightweight executable called a container. Considerably more information is available about that topic than the concept I asked about.

Therefore, the AI jumped to conclusions using the limited information on each data container concept. It mashed them together and presented that as the answer. Unfortunately, it’s flat-out wrong.

Interestingly, however, I see an immediate opportunity.

Opportunity for jumping to conclusions

If no human intelligence exists to feed generative AI tools, they’ll jump to false conclusions. But since B2B marketers have been jumping to conclusions for years, you use AI’s weakness to your advantage.

If you’re in a niche B2B market, AI’s false conclusions can propel or at least inspire you to find your version of “inbound marketing.” You can create content that defines (or redefines) the industry – the information that separates and sets new standards for your solutions to problems. You can more easily set the “right answer” for what generative AI should deliver.

You can teach your audiences as you train the machine.

This opportunity requires a renewed focus and big-time human output of thought leadership, content, and idea messaging. It also means you cannot lean on the traditional ways of defining what you do. You should learn what and how AI thinks of your industry, your approach, and your terms of art. See what your buyers can experience through these AI tools.

If HubSpot had focused on “article marketing” as its core thought leadership idea, it might never have differentiated. Instead, it stumbled (brilliantly, I might add) into a redefinition of “article marketing” and created a concept that became the standard answer.

That’s an opportunity for all businesses, but it’s a uniquely immediate opportunity for those of you in a niche business.

Did I just jump to conclusions?

You bet I did.

It’s your story. Tell it well.

Cover image by Joseph Kalinowski/Content Marketing Institute