The AI Visibility Spam Emails Are Wrong. Here's Why.

Nearly every morning, my inbox contains another email warning me that KaiMail, or Kafkai or PuchiDen or even my neighbours cat is about to become invisible.

The pitch is always the same. Someone has built a tool that queries ChatGPT, Perplexity, or Gemini to see whether our brand appears when users ask "What's the best tool for X?" If we are not in the answer, the email insists, we have already lost the buyer. No click. No tracking. No retargeting. Just oblivion.

This week it was SearchXXXX. Last week it was GPTXXXXXX. The names change but the script does not: AI is the new front door, being recommended is not optional, and for a modest fee they will fix it.

I have written about this topic before — in my article on why SEO is not dead but GEO is not optional and in the piece dismantling the Google Zero myth. The spam keeps coming anyway. So let me be direct: the claims in these emails do not hold up to scrutiny. And I am not the only one saying so.

The GEO Grift Is Real

So I follow Lily Ray who is a voice of reason within this sea of hallucinations. She recently founded the SEO and AI Search consultancy Algorythmic after fifteen years in the industry and published a detailed reflection on SEO and AI search earlier this year. She calls the wave of opportunistic GEO services exactly what it is: a grift.

Ray points out that many of the tactics now being sold as "Generative Engine Optimization" are simply core SEO approaches repackaged under a new acronym. Structured data, clear headings, concise language that answers real questions, keyword-optimized URLs — these are not discoveries. They are the fundamentals that SEO practitioners have been implementing for years, now being presented as revolutionary AI-specific strategies.

As she puts it, some of these recommendations are so familiar that they amount to proclaiming fire as a new invention. The only difference is the price tag and the urgency.

You Cannot Reliably Track What AI Says

The spam emails all hinge on a single premise: that AI visibility is measurable, trackable, and therefore optimizable. The problem is that the research says otherwise.

Ray highlights a critical limitation that these tools never mention: LLM responses are non-deterministic and highly personalized. They vary based on conversation history, memory settings, geography, and the specific phrasing of the prompt. Unlike traditional search engine results pages, there is no standardized result set to monitor.

This means when SearchXXXX or GPTXXXXXX claims they have "tested 20 real buyer queries" and found you missing, they are testing a snapshot of one possible response in one possible context. Change the wording slightly, use a different account, or wait an hour, and the answer may shift entirely. The data is directional at best, and presenting it as a definitive visibility score is misleading.

Dan Petrovic, another researcher Ray cites, has argued that GPT-5 operates as a reasoning engine that offloads factual knowledge to external search tools — which makes traditional SEO infrastructure more irreplaceable, not less. Google's AI operates on a grounding budget of roughly 2,000 words, extracting small sequential slices from content. Sometimes it uses as little as 13% of a long-form article.

Well, what does this all mean?

It means that if your content is not crawlable, indexable, and authoritative in the first place, no amount of "AI optimization" will help you appear in generated answers.

The Tactics Being Pushed Are Dangerous

What makes these emails particularly troubling is the advice that often comes bundled with the audit. Ray documents several dangerous GEO tactics that have been circulating:

• Using AI to generate thousands of new pages for "topical coverage"
• Buying aged Reddit accounts to seed "organic" recommendations
• Hiding LLM instructions via white text on white backgrounds
• Creating scaled-down page versions specifically for AI crawlers
• Advising clients to "write for bots, not for humans"

As Ray noted on X (you'll need to login to read it), some of the people promoting these tactics have never faced the wrath of Google's Helpful Content Update. That update specifically targeted content created primarily for search engines rather than people. The outcome is all too often predictable and devastating: a short-term visibility spike followed by a manual action or algorithmic suppression that takes months to recover from.

I covered the same risk in Part 1 of the AJSA seminar report: Google now evaluates page-level user satisfaction in real time. If users click through and bounce back to the search results, you drop. The era of gaming rankings with manufactured content is over, and the same mechanisms that detect low-quality SEO will detect low-quality GEO.

What Actually Works

Here is the part the spam emails get right, buried under the scare tactics: visibility in AI-generated answers does matter. But the way to achieve it is not through specialized GEO services. It is through the same discipline that has always worked.

A solid, cohesive SEO, social media, and digital PR strategy is, as Ray puts it, "by far the most effective way to capture visibility in AI search." The research backs this up. Princeton University studies found that content with clear upfront answers, structured formatting, data-backed claims, and credible citations performs better in generative engines. These are not exotic techniques. They are what good SEO and content strategy have always emphasized.

In Part 2 of the AJSA seminar report, I broke down the practical steps: answer questions clearly upfront, back claims with data, structure content with FAQ-style sections, and focus on first-hand research that is worth citing. Mr. Tsuji's decade of Japanese search data, which I referenced in the SEO Is Not Dead article, arrives at the same destination from a different angle: Google now measures post-click satisfaction and reflects it directly in rankings. Help the user, and both human readers and AI engines will reward you.

The Irony

There is a particular irony in receiving unsolicited emails claiming that AI search is making organic visibility impossible — while those same emails arrive through a channel that depends entirely on organic visibility to exist. These tools can afford to send cold outreach because someone, somewhere, is still searching, clicking, and reading.

The deeper irony is that the most effective way to show up in AI answers is to be genuinely worth citing. That requires original research, consistent quality, earned authority, and a reputation that accumulates over years. It cannot be purchased in a quarterly subscription to an AI visibility dashboard.

AI search is evolving. The zero-click era I wrote about is real, and the structural shifts are significant. But the water has not dried up. As Mr. Tsuji's data showed, 99% of businesses can afford to take a measured approach. The panic is not just premature — it is profitable, but only for the people selling the panic.

Your brand is not invisible because an AI chatbot failed to mention it in one snapshot query. It is invisible if your content offers nothing worth citing in the first place. I've also spoken about this from a branding angle in an intimate gathering of small business owners in Melaka.

Make your content worth reading and citing, and the rest follows.