Don’t Blame AI for Generic B2B Content Marketing

Remember playing with paints as a kid and mixing every color together? You expected a rainbow, but you ended up with a murky brown or dark gray.

That’s the same disappointing feeling you get when AI pumps out another piece of generic content. But what did we expect?

Large language models are trained on everything from Shakespeare to forgotten BuzzFeed listicles. When you blend everything together, it’s no wonder you get writing that’s average. Along the way, generative AI also learned how boring B2B blog posts are supposed to sound.

And therein lies the rub. We had a problem with generic B2B content before Claude and ChatGPT hit the scene. Now it’s getting worse. On the bright side, there’s never been a better time to start standing out.

Average in = average out: The real reason AI content is generic

Along with the works of Ernest Hemingway and Emily Dickinson, LLMs have an untold amount of B2B marketing articles in their training data. When you ask AI to write a blog post, it’s not going to reference Hunter S. Thompson; it will try to replicate other blog posts.

Unfortunately, most B2B blog posts are painfully generic. As content marketers, we have no one to blame for that but ourselves.

We played it safe, and now AI plays it safe. We wrote bland corporate jargon, and now it shows up in AI-generated copy. We overused the rule of three in writing, and now… You get the point.

An LLM uses pattern recognition to generate text. That’s why some people call it “spicy autocomplete.” The model is predicting likely language based on patterns it has seen before. AI refers to its training data to help it decide the next most likely word to answer your question or complete your article.

But the next most likely word isn’t necessarily the next best word. Most likely is average.

Where did those telltale signs of AI come from?

Plenty of people think humans are starting to write like AI. What’s more likely is that AI has been writing like us, and we noticed mediocre habits reflected back at us.

Let’s not open the debate on em dashes for the umpteenth time. Instead, let’s recognize that some of our cherished writing crutches and favorite phrases made their way into the training data.

It’s easy to mock AI-generated writing that starts with “In today’s fast-paced digital landscape.” But take it from someone who’s written and edited plenty of marketing and tech-related blog posts — there was a time a lot of us thought it made for a solid opening line.

Phrases like that are lazy, sure. There are also patterns AI picked up on that are useful and powerful. Sadly, we’re now questioning ourselves any time a pattern like “It’s not just X, it’s Y” shows up in our own work.

That could be effective. But since AI overuses antithesis as a rhetorical device, it certainly feels like you’re doing something wrong.

Take the word “leverage” as an example. When you see it in B2B content, you’ll probably jump to the conclusion that it’s AI-generated. 

Do you know who’s used the word leverage a lot? Seth Godin — the legendary entrepreneur, best-selling author, and marketing sage. Here’s a partial screenshot of site search results for “leverage” on Seth’s blog:

For context, when you publish a blog post every day for almost 25 years (as Seth has), there are a lot of opportunities to use the word leverage.

Yet there’s a huge difference between Seth’s books and blog posts and generic AI content. When Seth uses the word leverage, it has meaning and purpose. It’s a key concept in the ideas he’s trying to spread. 

When AI uses leverage, the best-case scenario is that it’s trying to sound like Seth Godin.

How B2B content became an echo chamber

Content marketing isn’t easy. To make it less challenging, plenty of writers looked for shortcuts.

Unfortunately, when you take too many shortcuts, you can lose your way. 

For example, “research” for a mediocre content marketer often meant Googling what competitors wrote on a topic and reworking those blog posts into something new. But there’s a good chance brands in your space did the same thing.

Some marketing teams aggregated long lists of industry stats to earn links and traffic. Then they hoped other lazy content writers cited them as the source, which happened. To make things worse, those marketers often updated the year in the headline without updating findings.

In the name of SEO, content marketers also used optimization tools that compared their drafts to the top-ranking content for a target keyword. That encouraged more copying of existing content as well as bloated articles with high word counts that defined every term imaginable.

Content got further and further away from authoritative sources, expertise, and original thinking.

All this led to the so-called sea of sameness in B2B content marketing. And that sea is getting deeper as generic AI content gets published at scale.

Do better tools and prompts make AI content less generic?

You’ll hear that if AI outputs generic content, it’s because you gave it a generic prompt. That’s absolutely true. Ask any LLM to write a blog post on a random topic and it will generate something potentially usable — but also forgettable. In other words, it’s an average blog post.

Provide the AI with a detailed brief including the audience, the goal, a working title, an outline to follow, and the results will improve. Upload reference material or use a knowledge base with documents that give it product information and brand guidelines, and it gets even better. 

Still, while the output gets more specific and accurate, it may still read as generic content. Here’s why:

• Predictable language patterns still show up.
• It defaults to a safe position instead of taking a bold stance.
• It has no real life experience to draw from.

Your content can be on-brand and still sound generic. It can even target the right audience, but still fail to resonate.

I’ve done a lot of experimenting with AI and content generation. Beyond writing, I ask it to push back and pressure test my ideas. That gets frustrating.

There are times when AI tells me it’s going to help “sharpen” some copy, but to me, it feels duller instead. Ask it to help you “cut the fluff” and suddenly the personality you loved about the piece disappears.

If you use AI to assist with B2B content marketing, you should write a detailed prompt, give it a brief, and upload helpful reference material. Just make sure you stay in charge and push back when you’re unsatisfied.

Use human judgment and avoid letting AI lead the way

Tools are useful, and many AI tools are pulling off mind-blowing tasks. The risk is forgetting that they’re tools and letting AI become the boss. I think we should wait a few more years before we start working for AI and make sure it works for us.

Let’s say you use AI to create a project or custom solution that’s instructed to give you critical feedback on B2B content. That’s an excellent idea. But tread carefully. Don’t view the AI’s feedback and suggestions as definitive. Take what works and question what seems off.

You may find that an AI solution designed to offer a content critique does so no matter what. That’s because critiquing is its purpose. It has one job to do and it does it.

AI even ignores its own advice. Not long ago, ChatGPT told me to take the phrase “move the needle” out of some landing page copy. I explained that I understood it was cliche, but I had a reason to use it. The AI continued to warn me about it throughout the chat.

The next day, I wanted to brainstorm a headline for a new project. Its first, uninspiring suggestion contained “move the needle.”

Sometimes, you’ll find yourself writing about things in theory rather than practice. AI can be helpful when trying to get a grasp on a technical topic. So, when you lack experience or expertise, use AI to learn faster, and then write about it using your own voice.

Rely on your own taste and judgment when editing AI-generated content. That’s the best way to avoid being average and generic.

Let’s take charge and make things right

B2B content marketers helped make this mess of generic content, and AI just made it worse. But blaming AI isn’t a solution. We’re the ones using it, and we’re the ones who can make it better.

Think of this transition as an opportunity to reset. You don’t have to pick sides. This isn’t a binary choice between letting AI do everything or relying solely on your content team. 

Let AI help you generate first drafts that a talented human can mold into something better. Take the insights and experience of experts in your company and see what AI can do with that. Use it to hone your ideas, explore complex topics, and get you unstuck as you write. Just make sure the right humans stay in the loop.

At Blue Ivory Creative, we’re helping B2B SaaS brands take advantage of AI without losing the originality and real-world experience that comes from human-led content.

When asked, we work to turn AI-assisted drafts, expert interviews, product knowledge, and scattered internal ideas into content that sounds like it came from people who know what they’re talking about — because it did.

Keep the efficiency of AI without publishing content that sounds like everybody else’s.