Your rankings look stable, but your traffic isn’t behaving the way it used to, and it’s not obvious why. Here’s the real cause: search volume as a metric is losing predictive power, because a growing share of queries never lead to a click at all. This isn’t a “keyword research is dead” piece. It’s a straight answer on what’s still reliable and what you should be watching instead.
A query showing strong historical volume in your keyword tool doesn’t mean what it used to. SparkToro’s zero-click search research confirms what a lot of marketers have felt anecdotally for a while now: a substantial and growing share of Google searches end without a click to any website, because AI Overviews or the answer on the results page itself already satisfies the query. Volume tells you people are searching. It no longer tells you whether they’re clicking through to you.
The gap runs the other direction too. Plenty of high-value, newer conversational queries show thin or nonexistent historical volume in traditional tools, not because nobody’s asking them, but because the data hasn’t caught up to how people are phrasing questions now. This is part of what an AI-first SEO strategy actually has to account for. Reading and adapting to shifts like this, rather than reacting to them a year later, is exactly the kind of ongoing work our Authority Engine™ is built to handle for clients.
Search volume isn’t retired, it’s just not the whole picture anymore. Three signals are worth adding to your regular check-ins, and none of them require enterprise software to start tracking.
Citation frequency: how often AI answers actually reference your content when responding to relevant queries. This is closer to a trust signal than a ranking one, and it’s worth checking directly. Pick your 15-20 highest-value queries, run them through Google, and note whether an AI Overview shows up and whether you’re cited in it. Keep a simple running list, date, query, cited or not, so you can actually see movement over time instead of relying on a gut feeling that something changed. You don’t need special tooling to do this once a month, a spreadsheet works fine.
This connects directly to how generative engines decide what to cite in the first place, which is a separate but related shift in how content gets structured and evaluated, not just tracked.
AI Overview presence matters because Google’s two AI search surfaces work differently from each other, and that distinction trips a lot of people up. AI Overviews are the summary box baked into standard search results. AI Mode is a separate, conversational search experience Google runs through what’s called query fan-out, running multiple related searches behind the scenes to build one answer. Because the two systems research differently, they frequently cite different sources for the exact same question, independent research comparing AI Mode and AI Overview citations found meaningful overlap gaps between the two even on identical queries. Treating “I show up in an AI Overview” and “I’d show up in AI Mode too” as the same thing is a real, common mistake.
Assisted-conversion thinking rounds this out. AI answers increasingly function as an early-influence touchpoint rather than a last click, someone sees your brand referenced in an AI answer well before they ever visit your site directly. It won’t show up cleanly in last-click attribution, but it’s real, and it’s worth acknowledging even if you can’t perfectly measure it yet. This same logic is exactly why we point to a concrete example rather than a theory: the Court House Lawyers case study is a verified AI Mode citation for a real client, in a vertical where, at the time of writing this, almost nobody else in this market can point to the same proof point.
A query showing “0” in your keyword tool isn’t the same as zero demand. It usually means the phrasing is newer or more conversational than the tool’s historical data can capture yet, especially for the kind of multi-part questions people now type directly into AI-driven search instead of a three-word query.
That gap matters because it’s exactly the phrasing AI systems are built to answer. If your content only targets the tidy, high-volume version of a topic and ignores the messier conversational version, you’re invisible to a growing slice of how people are actually asking. The fix isn’t abandoning volume data, it’s treating a zero-volume result as “unverified,” not “worthless,” and building content around real questions you can observe directly from customer conversations, support tickets, or the actual language your clients use, rather than only what a keyword tool has historically logged. A plumbing business might see zero recorded volume for a phrase like “why is my water heater making a popping noise before I call someone,” yet that is precisely the kind of specific, real question a homeowner would type into an AI-driven search and precisely the kind of page that could get cited answering it.
None of this means your existing tools are obsolete. They’re incomplete on their own, which is a very different thing.
Local, geo-specific volume data still matters enormously, and it’s not theoretical for us. Every local landing page CFM builds depends on knowing real search volume for a specific city or neighborhood; the same local SEO priorities that separate a local strategy from a national one still hold, and that data doesn’t get replaced by anything AI-driven. CPC alignment for paid search still depends on the same traditional data. And for client or stakeholder reporting, traditional rank tracking and volume metrics remain auditable and easy to explain in a way that “citation frequency” isn’t yet for most audiences.
Traditional tools also still reward what they’ve always rewarded: comprehensive coverage of a topic rather than a single keyword-stuffed page. That’s not a coincidence; it’s the same instinct behind building topical authority, and it happens to be exactly what AI Mode rewards too: genuine depth across a subject rather than a page optimized for one phrase. As you build that depth out, especially if you’re using AI tools to help draft it, keeping the process ethical and transparent in how AI is used in your SEO content matters just as much as the research behind it, since AI systems increasingly evaluate whether content reads as genuinely authored versus assembled.
The practical version of all this is simpler than it sounds. Keep your traditional keyword tools for volume, difficulty scoring, and local data, that part of your process doesn’t need to change. Layer in a lightweight manual habit: check AI Overview presence and citation status for your priority queries periodically, not obsessively; monthly is enough for most businesses. Review what’s ranking and what’s stalled the same way you’d approach identifying and reviving content decay, since a page that’s quietly lost its citation is a version of the same problem as a page that’s quietly lost its ranking.
This only works as an ongoing habit, not a one-time audit. Search doesn’t hold still long enough for a single check to stay accurate for a year, which is exactly why AI SEO workflows split between automation and human judgment rather than defaulting entirely to either one. A spreadsheet and a monthly calendar reminder is a legitimate starting point. What it isn’t is a permanent solution once the list of queries you’re tracking grows past what one person can reasonably check by hand.
If you’re a single marketer or small business owner checking 15-20 queries a month by hand, this is genuinely doable without outside help. Block a couple hours, run the searches, note what’s changed since last time, and adjust accordingly
Where it gets harder is once you’re managing this across dozens of service lines, multiple locations, or alongside everything else that actually keeps content up to date and technically sound, at that point, manual monthly checks start slipping the same way any manual process does when attention shifts elsewhere. That’s the layer our Authority Engine™ is built to operate at continuously: keeping traditional keyword data, AI citation tracking, and content freshness moving together as one system instead of three separate things someone has to remember to check. If you’d rather have that built and maintained for you than run it yourself every month, that’s the difference between checking this occasionally and search-driven lead generation that keeps working in the background.
Not sure if your current keyword approach is still working the way it used to? A free website audit will show you what’s actually happening instead of leaving you to guess from a traffic dip you can’t explain.
© Curious Fortune Media – All Rights Reserved – Terms and Conditions – Privacy Policy – Accessibility Statement
