A lot of SEO teams aren’t questioning whether SEO matters. They’re questioning whether they can still see it working.

Traffic looks flatter. Rankings don’t correlate cleanly with sessions. Leadership gets reports that are technically defensible but don’t feel like the full story. And underneath all of it are the same concerns:

“Why are my site’s clicks down?” and “If clicks are down, is what we’re doing still working?”

Those concerns are reasonable. Clicks were the clearest proof point SEO had. When they softened, alarms went off — not because SEO strategy had failed, but because the signal everyone relied on stopped behaving the way it used to.

Here’s why clicks are no longer the primary signal: search got more efficient. Answers surface earlier, making confidence build and decisions start forming before anyone even clicks on your site. And increasingly, those decisions are forming outside of Google entirely — in AI assistants, answer engines, and LLM-powered interfaces that often never send a referral signal at all.

The result is a growing gap between influence and attribution. That gap is what makes measurement feel broken and why SEOs need to start measuring different outcomes.

Why clicks were the benchmark for SEO success

In a traditional search experience, the path was linear:

Search → results → click → traffic → conversion. 

Clicks were the bridge between visibility and business outcome, and because that bridge was consistent, they became shorthand for SEO success.

The clicks model: Rankings up, clicks up, performance up. Rankings down = something’s wrong.

That model worked well in a world where search results were gateways to information. To get an answer, you had to actually visit a site. Visibility and traffic moved together.

The environment the clicks model was built for no longer exists in the same way, and the pace of that change has accelerated sharply in the last two years.

Why visibility now happens before the click — and sometimes outside Google

Modern search surfaces answers, not just options. AI Overviews, rich results, and AI-assisted experiences let users orient themselves, compare options, and often decide before clicking through. Zero-click behavior isn’t a new observation — it’s been building for years. What’s new is the scale and the expanded surface area.

A growing share of discovery and research is now happening in ChatGPT, Perplexity, Claude, Gemini, and similar tools. For technical buyers, founders, CMOs, and researchers — the audiences most SEO programs are trying to reach — LLM-assisted research is becoming a primary behavior. 

LLM-assisted platforms don’t cleanly pass referral traffic. They surface brand names, cite sources, and shape perception without ever showing up in your analytics.

These changes don’t eliminate the click, but they do reposition it. Clicks now reflect later-stage intent. 

What clicks mean now

When someone finally clicks, it usually means they’ve already seen the brand somewhere, recognized the name, trusted the explanation, and decided they’re close enough to act. The click is the outcome of prior visibility, not the moment influence began. The user is usually validating a decision they’ve already started forming. 

This is where many teams often get tripped up. Measure only the final interaction, and you miss everything that created the conditions for it. By the time someone visits your site, SEO may have already done its job — quietly, repeatedly, across touchpoints that don’t appear in a single dashboard.

A significant share of SEO’s actual influence is happening upstream: shaping perception, building familiarity, and narrowing the consideration set.

That’s true within Google. It’s even more true when you factor in LLM surfaces where your brand may be mentioned, recommended, or cited without generating a single trackable visit.

What to measure instead

The goal isn’t to replace clicks with a new vanity metric and continue the same as before. Search has changed too much for that to work. The goal is to measure SEO in a way that reflects how influence actually works now — across Google, across AI platforms, and across the full arc of a buyer’s journey.

Start by separating three things that often get blended together:

  • Visibility — are we showing up where decisions are forming?
  • Recognition — are people remembering and returning?
  • Outcome — is this contributing to revenue?

Clicks mostly capture outcome behavior: a moment of action. Modern SEO creates value well before that moment, so measurement has to expand upstream without losing accountability.

Four categories of signals hold up well when the click is no longer guaranteed:

1. Directional visibility across meaningful questions

Rank tracking alone doesn’t capture visibility anymore, especially when results vary by feature type, intent, and interface. The better question: are we present for the questions that define our market?

Track coverage across the topics your buyers actually care about. Look for expansion over time — more appearances, more consistent inclusion, broader coverage as queries branch. What you’re really measuring is whether you’re becoming a reliable source in the conversation.

A note on impression data: Google Search Console gives you a directional read here, but impression data has had well-documented reliability issues. Treat it as a signal, not a source of truth, and weight it alongside other visibility indicators.

2. LLM visibility and brand mention tracking

This is the measurement category that many teams don’t have yet — and it’s the one that’s becoming urgent.

The question isn’t just “do we rank on Google.,” but “when someone asks ChatGPT, Perplexity, or Claude about our category, do we show up? Are we cited? Are we named as an option, a leader, an alternative?”

LLM visibility tracking involves running structured prompts — best-of lists, alternative-to queries, VS comparisons, category-defining questions — and recording which brands surface organically in responses. This isn’t scraping. It’s systematic prompt-based auditing that tells you whether your entity is being recognized by the models shaping buyer research.

The brands investing in this now are building a data advantage that will be very hard to close later. Most of their competitors have no idea what the models are saying about them.

Learn more about our C.L.A.R.I.T.Y framework for AI → 

3. Brand-led discovery and recall

If SEO is building familiarity without the click, that familiarity shows up downstream as branded search growth, more direct visits, and navigational behavior — people searching your name plus a product category, a comparison, pricing, or reviews.

Those patterns aren’t random. They’re evidence of recognition. Leadership doesn’t need to believe in invisible SEO. They just need to see that SEO is shaping what people look for next.

Branded search trend is one of the cleaner signals available for demonstrating upstream influence. It’s not perfect, but it moves in ways that correlate with content ecosystem health and entity strength — and it’s something most stakeholders intuitively understand.

4. Assisted performance across channels

SEO rarely closes the deal alone. It influences the deal. That influence shows up as improved conversion rates on paid traffic because the brand feels familiar, shorter sales cycles because prospects arrive more informed, higher lead quality because content pre-qualifies intent, and stronger performance on product and solution pages because the content ecosystem around them builds trust.

Measure SEO only by last-click conversions and you’re grading it on the final step while ignoring everything it did to create the conditions for conversion. Accountability doesn’t go away — it becomes more accurate.

 How measurement changes are connected to content systems and entity strength

Measuring SEO and building SEO can’t be separated. Strategy determines what is measurable and what is durable. Build a system designed to compound, and the signals will align over time. Publish in isolation, and measurement will always feels noisy because nothing reinforces anything else.

Visibility itself is changing because search and LLMs are increasingly evaluating understanding, not just pages. The models that generate AI Overviews and LLM responses are drawing on entity relationships, topical authority, and source credibility. They’re asking: does this brand have coherent, consistent expertise across this domain?

The work that compounds in this environment looks like this: content systems create consistent reinforcement across topics, entity strength increases recognition and contextual trust across both search and AI platforms, internal architecture makes that structure legible and scalable, and a well-developed topical footprint rewards brands as queries branch and expand.

When those pieces are in place, measurement becomes less fragile. You stop chasing isolated ranking spikes and start seeing system-level expansion — more inclusion in Google, more mentions in LLM responses, stronger downstream performance across channels.

Measurement mistakes that create false negatives

A lot of teams conclude SEO isn’t working when the real issue is how they’re reading the signals. Maintain clarity by avoiding these measurement mistakes:

1. Treating session declines as the whole story.

If sessions dip while visibility is stable or expanding, the conclusion shouldn’t be “SEO is down.” It should be that click behavior changed — which requires deeper inspection, not a strategy shift.

2. Using outdated baselines as the benchmark.

Comparing performance today to a search environment from three or four years ago guarantees constant disappointment. Benchmarks need to reflect current interface realities, including the growing share of research happening in LLM platforms that send no traffic at all.

3. Over-grading individual pages.

Systems compound. Single pages fluctuate. Over-focusing on one piece at a time creates a culture of short-term reactions instead of long-term compounding.

4. Ignoring time lag.

Brand discovery rarely converts on day one. If you expect immediate linear outcomes, you’ll consistently abandon strategies right before they start paying off.

5. Having no LLM visibility data.

Most teams are flying blind on what AI platforms are saying about their brand and category. That’s not a minor gap anymore. It’s a measurement blind spot covering an increasingly significant share of buyer research.

How to talk to leadership about changing your SEO measurements

Leadership doesn’t need more data. They need a model that matches what they’re observing.

When talking to leadership, acknowledge the shift plainly: clicks are changing because search — and AI tools — are answering more questions before the visit (and sometimes without a visit ever happening). 

Recenter on the actual objective: being visible and trusted in the moments people decide, wherever those moments occur. Introduce the measurement layers — visibility, recognition, outcomes — and make clear that clicks are part of outcomes, not the full picture.

Then, connect SEO to broader performance. When the SEO system is strong, you see lift across paid efficiency, conversion rate, and lead quality. When LLM visibility is strong, you see it in branded search growth, in prospects who arrive already oriented, in sales cycles that move faster.

Finally, set expectations confidently: some signals are immediate, others are directional, and the right question is whether you’re becoming more present and more trusted over time — on Google and in the AI platforms where your buyers are increasingly doing their research.

What good measurement looks like going forward

Good measurement doesn’t abandon clicks, it just puts them in their proper place.

To put it in a nutshell, here’s what to do. Track directional visibility across the questions that define your market. Build LLM visibility tracking into your measurement stack — know what AI platforms are saying about your brand, your competitors, and your category. Watch for brand-led discovery and recognition signals. Evaluate assisted performance across channels. Interpret outcomes on the right timeline. Avoid making strategy decisions based on any single metric.

The brands that win in this environment won’t be the ones optimizing for yesterday’s dashboards. They’ll be the ones who expanded their measurement to match where buyer research actually happens — and built systems designed to be visible, trusted, and cited across all of it.

Remember—clicks still matter. They’re just no longer the primary signal.

Need help building an SEO system and the measurement infrastructure to match? Let’s talk.