If your website leaves visitors guessing about your core services, your messaging needs an overhaul. I’ve spent years convincing clients of this and working with them to improve their copy. AI is quickly taking over that side of the desk—convincing people that vague messaging needs to be changed. In fact, AI might be more effective at it than I am.
My prediction? Prospects using AI in B2B vendor searches is what will finally kill marketing buzzwords, business jargon, and word salad. We’ve all seen the “innovative, best-in-class solution that helps companies work smarter” copy—it’s as empty as it is commonplace, and leaves people with no clear impression of, well, anything. If AI is the catalyst that forces brands to trade meaningless word salad for specific, authentic communication, I will be loudly cheering it on from the sidelines.
HOW DO AI ENGINES INTERPRET YOUR MESSAGING?
The influence of AI in B2B and B2C sales is changing so quickly that the explanation of how AI engines find you would have been drastically different only six months ago. To understand how AI is answering user questions today, we need to look at two specific mechanisms: RAG systems and entity resolution.
Let’s say someone looks up, “What are the best running shoes for someone with flat feet?” As we covered in our last blog on how to get shortlisted by AI, part of AI engines’ data comes from what they were trained on, and part comes from real-time discovery. This real-time portion is using a retrieval augmented generation (RAG) system. Unlike traditional search engines that rely on exact keywords, they employ semantic understanding. In other words, a RAG system understands that “flat feet” and “fallen arches,” while using different words, mean the same thing.
Advanced AI engines (i.e., the versions of ChatGPT and Gemini you’re currently using) also employ entity resolution when guiding users toward answers. These engines look for concepts (e.g., person, product) and build understanding by mapping the relationships between those conceptual entities. In other words, if it finds several sites that claim “ASICS Gel-Kayano is a good shoe for people with flat feet,” the AI builds a cognitive map that links “ASICS” and “flat feet” together.
It’s critical to understand that AI engines aren’t looking for a list of possible answers. They are doing the opposite, and are instead searching for the exact right answer. So, AI engines may tell you that ASICS and Brooks are your best bets. Meanwhile, Nike, a company 35x the size of Asics and Brooks combined, is left out of the conversation.
Nike likely makes several excellent shoes for those with fallen arches, but Nike sells itself as being “for all runners.” The problem for Nike is that it misses the specific mapping. In the age of personalization, nobody is searching for “What are the best shoes for all runners?”
YOU CAN’T BE EVERYTHING TO EVERYONE
While that’s a B2C example, the implications for B2B are even more significant. It means that your ideal customer profile (ICP) is no longer just a guideline—it’s one of the most valuable marketing assets you have. If your messaging doesn’t target your ICP as clearly and accurately as possible, AI engines won’t just rank you lower—they’ll act as if you don’t exist at all.
As AI engines become more widely used, searches will become more specific. Why search for the “best CRM” when you can search for the “best CRM for a small startup with under 10 employees about to hire its first non-founder salesperson”? If your copy only says “best CRM,” you’ve already lost that lead.
It can be hard to limit your prospects when you’re hungry for growth. Especially when just starting out, you want to say yes to everything. But “limiting” your audience is exactly how you expand your reach. Here are a few good reasons:
- You can’t be everything to everyone. Your organization has strengths and weaknesses like everyone else’s. To be successful and scale, focus on what you’re best at.
- Customization becomes a time suck. Can you have different landing pages for various sectors? Of course. But if you have to reinvent your messaging and make a custom deck for every new potential client, it’s time to refocus your ICP.
- AI discoverability will become increasingly crucial. AI engines are curators, not librarians. If you aren’t specific to a certain vertical, a certain size company, or a certain use case, AI engines won’t present you as an option.
AI SEARCH KILLED THE MARKETING FLUFF
This brings us back to entity resolution and relationship mapping. If you’re trying to be everything to everyone, AI can’t identify a clearly defined entity. Not only will AI not map you to anything, but it will also be skeptical of your entire digital footprint. The statement “an innovative, best-in-class solution that helps companies work smarter” has zero data points when we break it down into defined entities. What kind of solution? What kind of company? Smarter how?
Strategi co-founder Adnan Baig has a great article in Medium explaining what happens in AI searches when company positioning is vague. Baig writes, “Vagueness introduces ambiguity, and ambiguity creates risk. AI systems are optimized to minimize the chance of being obviously wrong. If a brand’s positioning cannot be mapped clearly to a category, use case, or function, including it in an answer becomes risky. Risk leads to exclusion, not debate.”
Wait a minute. AI engines aren’t just looking at your website. In fact, to ensure the most accurate answer, they go to many different sites and especially factor in reviews on sites like G2 and even recommendations on sites like Reddit. So, even if Nike is messaging to every runner, couldn’t Nike come up in a recommendation for people with flat feet on another site?
The issue is that AI engines prioritize consistency. When third-party reviewers are more specific than your brand messaging, AI systems notice the disparity. As Baig explains it, “When their language diverges significantly from the brand’s own positioning, the system notices the mismatch. The brand’s description is discounted in favor of more concrete external explanations.”
There is one area where Baig and I disagree. He suggests that vague positioning used to work because humans could translate words less as definitions but more as signals for things like ambition or design. He claimed that these phrases could “feel impressive to a human reader.” I’d argue that these vague phrases never worked, and that people—on both sides of the messaging—were just too embarrassed to admit, “I don’t understand what these words are supposed to mean.”
Regardless of why it persisted, the era of marketing fluff is over. I don’t have to spend my time convincing people that they need clearer messaging anymore—AI is going to do it for me. It truly is “an innovative, best-in-class solution that helps companies work smarter.”
Or something like that.
Do you need a human to assess your sales messaging? Reach out to us at mastery@maestrogroup.co.