Why Ranking Alone No Longer Tells You Which Keywords Matter

You've done everything right: you picked keywords with solid search volume, you rank on page one — and yet traffic keeps slipping. The problem isn't your execution. It's that the yardstick itself has changed.

Traditional keyword research assumes a simple equation: high rankings equal visibility. In the era of AI answer engines, that equation is broken. According to an Ahrefs analysis from August 2025, only 12% of URLs cited by ChatGPT, Perplexity, and Copilot rank in Google's top 10. The overlap between ranking and being quoted by AI is far smaller than most businesses assume.

The gap runs in the other direction, too. An Onely study of 25,000 ecommerce queries found that 80% of sources cited in Google AI Overviews don't rank organically for the query at all — and even a top-three organic position yields only an 8% chance of being cited. As Onely's founder Bartosz Góralewicz puts it, "Google is pulling in sites beyond the usual top results."

Meanwhile, the clicks themselves are disappearing. A 2025 Semrush study shows zero-click searches have climbed to 58.5% in the US and 59.7% in the EU — up from roughly 25% just five years ago. When an AI summary appears, only 8% of users click a traditional link, compared to 15% without one, according to Pew Research data.

This shift changes what "the right keyword" even means:

• Volume no longer predicts visibility. Over 68% of queries triggering AI Overviews get 100 or fewer monthly searches — the long tail is now the AI battleground.
• Intent determines exposure. Over 88% of AI Overview-triggering queries are informational, not commercial.
• Ranking reports miss the real question. As practitioners at Windmill Strategy frame it: "Are we being referenced in Google's AI-generated answers?"
• Citation is the new prize. Cited pages earn 35% more organic clicks than non-cited competitors, per Seer Interactive research.

None of this means abandoning conventional SEO. Generative engine optimization and traditional SEO overlap by an estimated 90–95%, according to industry analysis — strong rankings, useful content, and authority still form the evidence layer AI systems draw from. But keyword selection must now account for answer engines, not just blue links.

That is exactly why AI SEO Consultants builds keyword research around both tracks: the commercial terms that still drive traditional rankings, and the question-based queries where AI engines decide which businesses get quoted. The goal is no longer just ranking on page one — it's becoming the source an AI engine trusts enough to cite.

The Two Keyword Types You Need: Commercial Terms and Informational Questions

Keyword research used to mean chasing volume. Today, the data shows that strategy misses where AI search actually lives. Over 88% of queries triggering AI Overviews are informational, and more than 68% of those terms get 100 monthly searches or fewer. That's not a long-tail opportunity — it's the primary battleground for AI visibility.

Meanwhile, commercial keywords tell a different story. Research from Semrush shows that 95% of AI Overview-triggering keywords show no ads or minimal CPC, and terms with a CPC above $2 remain largely unaffected by AI Overviews. In other words, the keywords that drive revenue still play by traditional SEO rules, while the questions your prospects ask before they're ready to buy have moved into a new citation economy.

This split demands a dual-track approach. AI SEO Consultants structures keyword lists around two distinct intent categories rather than volume tiers alone:

• Commercial terms — service keywords, "near me" queries, and high-CPC phrases where organic rankings and Google Business Profile signals still drive leads directly
• Informational questions — the specific, low-volume "how," "why," and "what if" queries that trigger AI Overviews and answer-engine citations

The research backs this segmentation. Informational keywords drive roughly 90% of AI Overview triggers, while commercial intent accounts for only 6–8%. Yet keywords with commercial value retain their traditional SEO worth because AI Overviews rarely appear where ads dominate. A Windmill Strategy analysis notes that GEO and SEO share an estimated 90–95% overlap — the fundamentals haven't changed, but the informational layer now requires a citation-first content structure.

Treating every keyword the same way guarantees wasted effort. The winning lists separate commercial targets that need ranking power from informational targets that need citable depth — statistics, expert quotes, and structured answers that answer engines can extract and trust.

A Step-by-Step Keyword Research Process for AI Visibility

Keyword research for AI search isn't guesswork — it's a repeatable process that combines direct platform testing with traditional tools. Here's the exact workflow consultants use to build keyword lists that earn visibility in both Google and AI answer engines.

Step 1: Query the AI platforms directly. Type your target questions into ChatGPT, Perplexity, Gemini, and Google AI Mode, then record which sources get cited. This matters because ranking and citation are now disconnected — an analysis of 25,000 ecommerce queries found that 80% of sources cited in AI Overviews don't rank organically for the query at all. The only reliable way to see who AI trusts in your niche is to ask it.

Step 2: Mine traditional tools for citation gaps. Take what you learned from the AI platforms into Semrush, Ahrefs, and Google Search Console. Look for questions where the currently cited sources are thin, outdated, or generic — those are openings where you can provide a better, more citable answer. According to research on AI search behavior, this reverse-engineering approach is now central to modern keyword research, alongside tracking AI Overview appearance rates as a new success metric.

Step 3: Apply the fan-out framework. Don't stop at one version of a question. Expand each core topic across five query categories:

• Broad educational questions ("how does tankless water heater installation work")
• Comparative questions ("tankless vs. traditional water heaters")
• Technical specification questions ("what size tankless unit for a 3-bathroom home")
• Application-specific questions ("tankless heaters for cold climates")
• Operational and cost questions ("tankless water heater installation cost")

This structure, outlined in practitioner guidance on GEO, mirrors how real users phrase prompts to AI engines — as full questions, not keyword fragments.

Step 4: Prioritize informational, question-format queries. The data is unambiguous: over 88% of queries that trigger AI Overviews are informational, and more than 68% of those terms get 100 or fewer monthly searches, per Semrush's AI search data. Low volume no longer means low value — it means low competition for an AI citation. Keep higher-CPC commercial keywords on a separate traditional SEO track, since they remain largely unaffected by AI Overviews.

Step 5: Re-check your list regularly. Treat your keyword list as a living document, not a one-time deliverable. AI Overview presence more than doubled after Google's March 2025 Core Update, according to a 30,000-keyword tracking study, so a query that was safe six months ago may now be answered entirely by AI.

This is the process AI SEO Consultants runs for clients — querying answer engines, mapping citation gaps, and refreshing keyword targets as AI behavior shifts. The output isn't a static spreadsheet; it's a prioritized list of questions your business is positioned to answer better than anyone currently being cited.

What Makes Content Worth Citing (So Your Keywords Actually Pay Off)

Keyword targeting only works if the content earns citations. AI systems don't reward keyword density — they reuse evidence. A Princeton and Georgia Tech study found that adding statistics boosts AI visibility by roughly 41%, while expert quotes lift it by about 28%. Verifiable author authority can increase citation likelihood by up to 340%. These aren't formatting tricks; they're the evidence layer that answer engines cite.

• Statistics and original data points that support claims
• Expert quotes with named credentials and affiliations
• Author bios that demonstrate topical authority
• Structured formats — FAQs, tables, bulleted lists, clear headings

78% of AI Overviews feature lists, and Google typically cites six to fourteen sources per response. Content built for semantic depth and snippable answers outperforms keyword-stuffed pages every time. The Berkeley California Management Review puts it plainly: generative engines evaluate semantic depth and quality, not keyword density. Keyword stuffing is obsolete.

AI SEO Consultants builds this evidence layer into every piece of content we produce — structured answers, cited data, and authority signals that AI systems recognize and reuse. The goal isn't just ranking. It's becoming the source an AI engine trusts enough to quote.

Getting an Actionable Keyword List Without Managing the Tools Yourself

Most owner-operators in the trades and professional services don't have an extra ten hours a week to query ChatGPT, Perplexity, and Gemini, cross-check which sources get cited, and then monitor whether Google's AI Overviews shifted after the latest core update. Yet that's exactly what modern keyword research demands.

The workload is real. AI Overview presence more than doubled in early 2025, jumping from 6.49% of queries in January to 13.14% by March, so a keyword list built once and shelved goes stale fast. Effective targeting now means treating your list as a living document — re-testing queries against AI platforms, tracking citation overlap, and watching which of your terms start triggering AI-generated answers.

This is where a managed consultancy changes the equation. Instead of handing you a dashboard and a login, a done-for-you approach assembles three evidence streams into one prioritized, actionable list:

• Traditional keyword data — search volume, difficulty, and intent, segmented so commercial terms with CPC above $2 keep their traditional SEO value, since Semrush's AI statistics research shows those keywords remain mostly unaffected by AI Overviews.
• AI platform testing — querying ChatGPT, Perplexity, and Google's AI Mode directly to see which sources get cited, then identifying gaps where your business can provide a better, more citable answer, a process outlined in PushLeads' analysis of AI search.
• Local SEO signals — Google Business Profile optimization, review velocity, and citation consistency, because AI systems pull heavily from this local evidence layer when recommending service providers.

The output isn't a spreadsheet dump. It's a dual-track list: commercial keywords that still convert through classic rankings, plus question-based informational queries formatted the way AI engines prefer. That distinction matters, because over 88% of queries triggering AI Overviews are informational, and the majority get fewer than 100 monthly searches — exactly the long-tail questions a plumber, lawyer, or accountant's customers actually ask.

At AI SEO Consultants, the engagement runs month-to-month with plain-language reporting that shows which keywords you target, where you appear, and how visibility shifts over time. You own everything — the website, the content, the data, the domain — so nothing is held hostage if you ever walk away. And because no one controls what ChatGPT or Google's AI Overviews choose to cite, reporting stays directional and honest rather than dressed up with guarantees no one can make.

The practical starting point is knowing where you stand today. A free AI SEO Visibility Report shows whether your business currently surfaces in Google, Google Maps, AI Overviews, and the major answer engines — and which keyword opportunities you're missing. From there, the list builds itself around evidence, not guesswork.

Be the business AI recommends. Request your free AI SEO Visibility Report and get a keyword strategy built for how search actually works now — done for you, reported clearly, and owned entirely by you.