Memorandum
- From
- Derek Weston via MarTech
- Date
- Filed
- Business·5 min to read
- Re
AI tools will define a vendor's weaknesses if companies stay vague
ReAI tools will define a vendor's weaknesses if companies stay vague
Marketing agency Intercept argues B2B companies that avoid stating who their products are not for leave AI systems to fill the gap, often with inaccurate or unfavorable conclusions.
B2B companies rarely tell prospects they are a poor fit, but a marketing executive warns that this habit is becoming costly as buyers turn to artificial intelligence to research vendors. Andrew Au, managing partner at the B2B agency Intercept, argues that companies need to publish clearer boundaries describing who their products serve, where they fall short, and when another option makes more sense. Otherwise, he says, AI systems will fill in those blanks on their own.
Most product pages are built to keep as many buyers as possible in play, with copy that promises the product works for enterprises, growing companies, sophisticated teams and beginners alike. Au said the result is that the buyer, or the AI tool the buyer is using, has to work out whether the product fits. "What we all generate is how great we are and how amazing we are and how perfect we are," Au said. "So nobody talks about where we're potentially not the right fit."
Intercept and its Watchtower audience-intelligence platform analyzed about 2 million B2B conversations, according to Au, and found that technical and nontechnical members of buying committees were more likely to use AI to find new or overlooked vendors. Qualitative interviews pointed the same way. Au recalled one buyer who said he starts with vendors he already knows and then uses AI to find niche specialists he has never heard of.
That shift helps smaller and less familiar vendors gain attention, but it also gives buyers a faster way to test weaknesses. "The thing that the buying committee is using AI for is actually to find vendor weaknesses because everybody says they're wonderful," Au said. When a vendor claims to fit every kind of buyer without explaining trade-offs, Au suggests, someone or something else may decide where the product does and does not belong.
Au said he has seen this happen in audits where there is not enough useful content for AI models to draw on. In one example, he quoted a model's output that said a client was not suitable for certain scenarios, which he called inaccurate. "That's just not true. But when you look at it, when there's no content to actually review or cite, it will make an inference," he said. In his view, that inference can amount to a hallucination.
The remedy, Au said, is giving AI systems better material. A product built for large, distributed teams can state that it does not offer a self-serve plan. Software designed for one industry can say plainly that it does not support adjacent use cases. An enterprise platform can explain that a small business probably does not need its complexity yet. "If your product is for enterprise, don't pretend that you're for small business or for a startup," Au said.
The content itself is straightforward, but getting a company to agree on it is harder. Au said product marketing, sales, customer success and product teams need to settle which buyers the product is not right for. That can mean acknowledging that a competitor is the better fit in some cases, or telling sales that a prospect should not remain in the pipeline simply because budget is attached. Au said these discussions are uncomfortable. "You want to protect your brand truth, right? And with AI and humans," he said. "It sounds like an obvious thing, but it is a very uncomfortable conversation that we have with our clients."
Au said the boundaries do not need to dominate the landing page. They can appear where they fit naturally, including product pages, pricing pages, FAQs, buyer guides and comparisons. He framed the goal not as a confession page but as making the limits visible where buyers are already looking.
Clear exclusions also serve demand generation. If AI recommends an enterprise accounting platform to a freelancer seeking a cheap self-service tool, the referral is of little use: the prospect arrives, sees the mismatch and leaves, or, if they proceed further, sales spends time qualifying a deal that never belonged in the funnel. Stating exclusions up front can help AI systems and buyers rule a product out before the click.
Au's broader point is that more AI visibility is not automatically better. Marketing operations teams, he suggested, should judge AI referral traffic by what happens after the visit. If referrals rise while qualification, sales acceptance or opportunity creation falls, the company is gaining visibility without gaining better demand. Au also said comparison content will need to become more candid, since vendor comparisons that acknowledge competitor strengths before declaring the vendor the best choice may no longer hold buyers' attention.
5
