The keywords that actually convert don't show up in any SEO tool.

The playbook that worked. For a while.

For years, a big part of our SEO work at Growtika was building what the industry calls "top of funnel" content. "What is zero trust." "Types of phishing attacks." "How SIEM works." We found writers who actually understood cybersecurity, not copywriters who Googled it for twenty minutes, so the content was genuinely useful. We always did BOFU too: solution pages, comparisons, use case content. And we mixed in niche technical articles that no generalist would touch. But TOFU was one of the reliable ways to show growth. That's where the volume was.

And it worked. Across our client portfolio we maintained about 90 awareness pages pulling around 100K visits a month. Clients saw traffic grow quarter over quarter. Some of those visitors converted. The strategy looked good in presentations because it was good. And there's a real reason companies chase traffic: brand awareness matters, and traffic is a ranking factor. A site with high relevant traffic is more likely to rank for other pages too, including the BOFU keywords that actually close deals. So it all served the same mission.

The process was what every serious SEO team does: open your keyword tool, filter by volume and difficulty, build a keyword list, map it to content types, produce. Every month, the same cycle. Why would you go for a keyword with zero search volume when another one shows 4,000? The informative keywords were known to convert less, but their value was in the ecosystem they created. It's a proven system.

But something kept nagging at me. The highest-traffic pages and the highest-converting pages were almost never the same pages. When I finally mapped it properly:

Real data from standard SEO tools, pulled this week. Every single keyword in the bottom group returns zero volume. The tools say nobody searches for them. But these are the exact phrases that come up on sales calls the week before someone signs a $200K contract. And the top group? KD 40-51. You're fighting Microsoft, CrowdStrike, and Gartner for months to rank for visitors who are mostly students and researchers. The bottom group has no competition because no tool tells anyone to build it.

Then AI accelerated the shift.

Google's AI Overviews started answering the exact questions those TOFU pages existed for. "What is SIEM?" gets synthesized right in search results now. No click needed. LLMs became the first place technical buyers check. A CISO doesn't Google "what is CASB" anymore. They ask ChatGPT. Three seconds, done.

Traffic to those 90 TOFU pages started dropping. And that's when the ecosystem logic breaks: if the TOFU traffic that was supposed to boost authority and help BOFU pages rank is no longer reliable, what happens to the whole funnel? The keywords that looked great in presentations, the ones with 4,000 monthly searches, suddenly weren't delivering the traffic that justified them. People just preferred to check ChatGPT and Claude rather than clicking through to a site.

That doesn't mean you stop building TOFU. You keep going. But you double down on what makes sense right now: solution pages, use cases, very high-intent BOFU keywords that help the target audience before buying decisions and when they have an actual issue to solve.

AI didn't break our strategy. It made the old playbook table stakes. The new edge lives in content no keyword tool can find.

The keywords that convert don't show up in any tool.

"How long does SIEM deployment actually take?" shows zero volume in every SEO tool. Zero. The tool says nobody searches for it. But it's the exact thing a VP of Security asks the week before signing a $200K contract. If you have content answering that with real deployment timelines from real projects, that page will outconvert any "what is SIEM" glossary page by an order of magnitude.

SEO tools aren't great below 100 volume anyway, so you have to trust your instincts. And that's exactly when SEO and GEO have to get involved not primarily with the tools, but with the team. The founders. The sales reps. You need to sit in on inner meetings, listen to sales calls, understand the issues that lead people to the company. The stuff that only inside information can give you.

And here's what makes this a strategic shift: there are hundreds of these keywords hiding in every company's sales conversations. Each one has zero competition because no tool tells anyone to target it. "What is EDR" is KD 51, you're fighting Microsoft, CrowdStrike, and Gartner for position. "EDR rollout what goes wrong" is invisible to every keyword tool. You rank faster, for less effort, and the person reading it is actively evaluating a purchase.

That's when SEO starts to get genuinely interesting. You're not just a keyword researcher feeding a content machine. You're an investigator. You're extracting bullet points from internal meetings, decoding the language prospects use on calls, understanding pain points that no public data can surface. The SEO person needs to be in the room, or at least on the recording.

TOFU still matters. We still build it. It establishes topical authority and makes the BOFU content rank. But the energy equation has flipped. The zero-volume, zero-competition keywords that attract buyers are sitting right there in your sales team's call recordings. Nobody is building content for them.

SEO is going offline. I know how that sounds.

The web is flooding with AI-generated content sites. They publish anything. They link to anything. The signal-to-noise ratio of backlinks is getting worse. From what we're seeing, a mention in a real publication with real journalists is starting to carry more weight than links from sites that exist to sell links.

For LLMs the logic seems simpler. From our testing, links don't appear to be how a model decides what to recommend. But mentions in credible sources do seem to matter. When respected cybersecurity publications or conferences reference your product in context, that pattern shows up in how models talk about the category.

I've been doing B2B SEO for 10+ years. I never expected "go talk to people in your industry" to become SEO-adjacent advice. But here we are.

The old approach of opening your keyword tool, building a list based on volume, and fighting for competitive rankings is becoming table stakes. Everyone can do it. AI makes it even easier. And the keywords with volume are the hardest to rank for. You're spending the most energy on the lowest-converting content.

The edge is in the questions your sales team hears that nobody else is targeting. Zero competition. Higher conversion. And there are hundreds of them in every company's sales conversations, support tickets, and Slack channels. Each one is a page your competitors will never build because no tool told them to.

We still build TOFU. It's the authority layer. But the content that moves pipeline comes from a source no tool can access: the conversations happening inside the company, one week before a buying decision.