How to Measure What Matters in the AI and Zero-Click Age
Every marketer in SaaS feels the pressure to do something about AI. Predictions of it replacing marketers – and knowledge workers at large – are everywhere, and your leadership team is probably asking what your AI strategy is. It feels like if you’re not overhauling your entire roadmap, you’re already behind.
But what if the hype is just… hype?
What if the biggest threats to your marketing performance measurement have nothing to do with AI? In a recent CaliberMind webinar, “A Marketer’s Survival Guide in the Age of Zero Click Everything,” SparkToro co-founder Rand Fishkin dropped a reality check we all need to hear. The data shows that while we’ve been distracted by the shiny new thing, the ground has been fundamentally shifting beneath our feet. It’s time to look past the noise and focus on the real challenge—the rise of “zero-click everything”, and what it means for marketing measurement.
Why AI Hasn’t Changed User Search Behavior (Yet)
Before you scramble to devise an AI/AEO strategy, take a breath. As Rand Fishkin explained, he is “deeply skeptical that AI has changed all that much” when it comes to actual user behavior.
The data backs this up.
- Search volume is stable: The number of Google searches performed by the average person in the US hasn’t gone down. People are not, in large numbers, replacing their search habits with AI chatbots.
- People are still browsing: They’re still on Reddit, Pinterest, and Facebook – and in relevant community Slack channels. The platforms we’ve been using for years aren’t dying off.
The one place AI has changed everything is in media hype. Slapping an “AI” label on something is a great way to get noticed – and drives better email open rates (this fad won’t last!), but it doesn’t mean your entire audience has changed how they find and buy products.
The Real Threat to Your Marketing Data: Zero-Click & Its Implications for Attribution
The intense focus on AI is distracting from two trends that are actively breaking our traditional measurement models.
- The Rise of “Zero-Click Everything” – Search engines (whether it’s Google, Perplexity or any LLM acting as one) are increasingly becoming walled gardens, answering user questions directly on the results page with featured snippets and AI overviews. Google is keeping people on Google. Same goes for any other platform. This means fewer clicks to your website, even if you rank #1. That referral traffic you’ve been reporting on for years is drying up, and it has little to do with ChatGPT. But what if most of that top of funnel traffic you used to report on is just “tire-kicker traffic” anyways, and would never convert?
- The Erosion of Organic Attribution – For any MOPs pro who have spent months trying to perfect their click-based attribution, this one hurts.
Tracking a user’s journey digital from start to finish has become harder and harder, especially if you don’t have specialized tools to help you out.
Here’s why:
- Zero-click content is prioritized by search/social platforms, with users less likely to leave social/search platforms to consume content.
- When users do navigate off platforms, many platforms are stripping referral data.
- Privacy blockers and new laws limit tracking.
- Users are spread across multiple devices.
Attribution will never be 100% perfect, for a litany of reasons. These factors affect everyone, but organizations without account identification in their toolbelt are especially at risk. Much of your search / social engagement can easily be missed if you rely solely on users identified via direct conversions. And while you are looking at your click tracking, did you forget the mountains of engagement data you already have in your MAP and your CRM and many other disjoin platforms that could probably tell you 70% of the story if you could bring it all together? It has never been just web anyways…
Rethinking Marketing KPIs: Why “Increasing Traffic” Is a Flawed Goal
So, if attribution is tricky and traffic is harder to get, what should we be measuring? Unfortunately, many teams are still stuck in the past. A HubSpot survey found that “increasing traffic” is still the #1 priority for most marketers.
As Fishkin put it during the webinar, this is a “bad priority”. It’s a flawed goal for two simple reasons:
- It’s harder than ever to get: With zero-click search on the rise, the pool of available referral traffic is shrinking. And more traffic never meant more conversions anyways.
- It’s a poor indicator of success: A buyer might interact with your brand 15 times—on Reddit, YouTube, a podcast, etc.—and only visit your website once to make the purchase. In that scenario, your traffic is way down, but your revenue is right where it needs to be. Did you fail? Or did the metric?
Tactical Guidance for Marketing Ops:
Instead of abandoning the traffic metric entirely, make it smarter. Stop reporting on total traffic and start segmenting it. Isolate and report on traffic from your Ideal Customer Profile (ICP) or target accounts. This “qualified traffic” is a far more accurate indicator of business impact and helps you educate leadership by shifting the conversation from sheer volume to actual value.
A sample CaliberMind report showing website traffic by account tier
A New Measurement Playbook: From SEO to AIO/AEO
So where do we go from here? The new strategy is “Search Everywhere Optimization,” which means acknowledging that search happens on Reddit, LinkedIn, and yes, even LLMs—not just Google.
For marketing ops, this means expanding your measurement toolkit. While the overall impact of AI on search may be overhyped, you can’t ignore it completely. Here’s how to start measuring your visibility in this new channel, based on tactical advice from a recent Hackernoon newsletter on AI Optimization (AIO) / Answer Engine Optimization (AEO).
- How to Measure Your “Share of AI Answer” (SAA): Treat your visibility in LLMs as a new KPI. Routinely ask models key questions like “Best [your category] for [your ICP]?” or “[Your product] vs [competitor]?” Track whether you are mentioned and how. You should also monitor your referral traffic from LLM tools in your web analytics.
- How to Create “Canonical Answer Pages” (CAPs): To improve your SAA, work with your content and web teams to create clean, quotable pages that are easy for models to ingest. This includes clear pricing tables, technical FAQs, and security policies.
- How to Add Machine-Friendly Context: Ensure your site uses technical SEO best practices like schema.org and clean sitemaps. LLMs love structured data.
A sample CaliberMind report of referral traffic from LLMs such as ChatGPT, Perplexity, Claude, and Gemini
Your Next Steps in Modern Marketing Analytics
Success in modern marketing operations requires adapting to a world with less web traffic data and switching our attention to bringing together all the other fragmented engagement data that we have plenty of – but never cared enough to join (or never had the ability to do so). We have to look past the AI hype and focus on what really matters – squeezing what we have out of the data that we have collected today… You’d be surprised how much data you have if you take the time to look. You may even have enough to help you make a better decision than having all of those irrelevant clicks that went away with zero-search. So, zero-click search might actually reduce the noise and make your job easier—who knew?!
The path forward is to shift from chasing vanity metrics to measuring influence, educating leadership with better data, and investing in long-term brand building over short-term “hacky baloney,” as Fishkin called it.
This is just the start of the conversation, of course. If you want to hear Rand Fishkin break all of this down himself, you can watch the full on-demand session. And if you’re committed to rethinking your analytics this year, we hope you’ll join us for the rest of the Hitchhiker’s Guide to Marketing Analytics series, airing live every Wednesday at 1 PM ET through October 15th.
