Zero-Click, AI Search, and the Last-Click Attribution Trap: What B2B Marketers Need to Know with Amanda Natividad 

A lot of marketers have made peace with zero-click search by now. Then AI search showed up and raised the stakes again.

Amanda Natividad has been tracking this shift since 2022, when the concept barely had a name. She's the chief evangelist at SparkToro, the audience research tool, and she's co-authored Zero-Click Marketing: How to Build Brands and Earn Customers on a Traffic-Starved Web with Rand Fishkin. 

This is Amanda's third time on the podcast. She first joined in 2022 to talk about the early case for zero-click content, then came back in 2024 as attribution and audience research became the bigger conversation. This time, she's arrived with sharper numbers and a harder edge to her argument: AI search isn't a new front in the zero-click problem, it might be the most extreme version of it we've seen yet. 

In this episode, Amy and Amanda get into why AI visibility is the metric marketers need to watch, what a ghost citation is and why B2B marketers need to be aware, and why last-click attribution keeps costing B2B teams their best customers.

Why AI Search Is Making Zero-Click Marketing More Extreme

Half of Google searches ended without a click when Amanda Natividad first joined the podcast back in 2022. At the time, it felt like a huge shift. Four years later, that number has climbed to 68%, and zero-click search isn't a concept to be aware of anymore, it’s simply how marketing works.

AI has pushed things even further.

Ask ChatGPT a question and there's a good chance you'll get the answer without leaving the chat. Google still owns most searches. But it no longer owns the entire search journey. People search on Amazon while they shop, on social while they scroll, and inside Perplexity, ChatGPT, Claude or your LLM of choice when they need a quick answer. Bottom line, search has spread out.

For B2B marketers, that's created a problem hiding in plain sight. AI tools like Claude, Perplexity and Gemini rarely pass referral data, so someone who finds you through an AI recommendation often shows up in Analytics as direct traffic. You could be getting AI traffic today without ever realizing it.

That also makes last-click attribution less reliable than it used to be. If AI-driven visits are showing up as direct traffic, your reporting is already missing part of the story.

Which is a shame, because it's often some of your most valuable traffic.

Amanda shared research showing AI-referred visitors can convert at up to 10 times the rate of an average referral. The volume might be smaller, but the intent is often much higher.

That's why more marketers are paying attention to AI visibility, not just click-through rates. Tools like HubSpot's AEO Grader and Gumshoe.ai are finally giving teams a way to see whether they're showing up in AI-generated answers instead of simply hoping they are.

Also check out How to Win at Gen AI Search With Christopher Penn.

Ghost Citations Are the GEO Problem SEO Never Had

Amanda Natividad uses the term ghost citations, a phrase she credits to Kevin Indig, to describe a quirk of AI search that's becoming hard to ignore. It's when an LLM uses your research, your data, or your original thinking without telling anyone where it came from.

Take the 68% zero-click stat from SparkToro's research. Ask ChatGPT about zero-click search, and there's a good chance you'll get the number, but SparkToro might not get credit. That's a ghost citation. As a B2B marketer, it’s pretty frustrating!

It's a bit of a strange one because we're used to thinking about citations as links, names, and clear attribution. AI doesn't always work like that. Your work can shape the answer without your brand ever appearing alongside it.

Regular citations still matter in GEO as they do in SEO.

You might get a mention, a link, or your name attached to the idea. Ghost citations sit somewhere in between. They're not the outcome you're aiming for, but Amanda's point is that they're still a sign your content is influencing AI-generated answers.

A subtle shift, yes, but it changes what "getting credit" looks like in AI search. You can publish original research, introduce a new idea, and watch it spread through AI-generated answers without your name always popping up. It's a useful reminder that attribution isn't just a traffic problem. Just as last-click attribution can't capture every touchpoint in a buying journey, AI citations don't always reveal where an idea first came from.

How to Improve Your Visibility in AI Search

Retrievability and extractability sound more intimidating than they are at face value. In fact, most B2B companies are already doing a decent job of both. If an LLM can find your website and read what's on the page, including the transcript beneath your videos, you're probably in good shape.

The harder part comes next. Amanda describes credibility and public evidence as the two things you're never quite done working on.

Credibility is about whether you're known for a topic.

Public evidence is about proving it. Basically, it’s content that gives an LLM something concrete to point to; it could be case studies, customer stories, or testimonials.

It's not the flashiest work. You're probably not going to get hundreds of comments because you published another case study.

But that's not really the point.

When someone asks an LLM a high-intent question, it's already looking for evidence. If you haven't published it by then, you've missed your chance. It's a different way of thinking about search, too. Instead of asking, Which keywords do we want to rank for? Amanda suggests starting with a different question:

What would someone ask an LLM when they're genuinely looking for a solution? The answer makes it a lot easier to find, and you've done a lot of the heavy lifting GEO asks of you.

How AI Search Challenges Last-Click Attribution

One of the biggest differences between AI search and traditional search is how it connects people with products. Amanda saw that firsthand while building the website for Zero Click Marketing. She needed a contact form, asked Claude for a recommendation, and it pointed her to Formspree. 

So far, nothing unusual. It did all the things that she wanted it to do, and was simple to install.

Then she met Formspree's founder.

He asked if she'd ever seen the company's homepage. She hadn't. When he pulled it up, it looked like it had been built entirely for developers, code snippets in the hero image, developer-first messaging, the works. If she'd landed on that page through Google, she probably would've assumed it wasn't for her.

Turns out, she was exactly the kind of customer Formspree wanted. Her needs were low volume, just a handful of contacts a month, well within the free tier. And she was happy to pay once she outgrew it. AI didn't care that the homepage looked like it was written for developers. It cared that Amanda needed a contact form.

That's the shift. It's also one more reason last-click attribution can't tell the whole story.

For years, marketers have spent a lot of time defining audiences and building personas. AI search seems to start somewhere by asking, what is this person trying to do?

Why Last-Click Attribution Misses More Than You Think

Conversion windows are a bit of a made-up rule.

Some platforms stop counting after seven days. Others give you thirty. Neither number has much to do with how people buy. It's just where the tracking stops.

The average buying journey is closer to 40 days.

Which means some of your best customers, the ones who spend more time researching and often end up being your highest-value buyers, don't make it into the report. They didn't disappear. Your attribution window just closed before they converted.

Amanda isn't arguing that attribution is useless, quite the opposite. Keep measuring what you can, just don't stop there. She recommends adding incrementality tests into the mix. Switch a channel off for a while and see what happens when it's gone.

Dropbox did exactly that. In March 2026, its data science team paused mobile ads and search marketing for a month at a time. On paper, both channels looked healthy, but once the ads were switched off, the picture changed. Mobile was returning 76 cents for every dollar spent. Search marketing, 92 cents. Dropbox reallocated $25 million, lowered acquisition costs, and increased customer lifetime value.

Most teams aren't working with Dropbox-sized budgets, but the thinking still applies.

Amanda's favorite example is much simpler. Ask your CFO what podcasts they listen to. Ask whether they skip the ads. Then ask where they last discovered something they ended up buying. It's a quick reminder that not everything that influences a purchase shows up neatly in a last-click attribution report.

Public Mentions: A Better Attribution Metric Than Last-Click

Last-click attribution gives the credit to whatever happened right before someone converted. The problem is that's rarely where the buying journey started.

Amanda thinks there's a better signal to pay attention to: public mentions.

Dashboards are useful, so are reports, but there’s something distinct about people talking about you.

It could be everything from a LinkedIn post to a tweet or someone simply recommending your company in a Slack group you're not even part of. People don't usually go out of their way to mention a product or agency unless it made enough of an impression to be worth sharing.

Amy's seen that play out with Content 10x, too. Plenty of new client conversations can be traced back to someone mentioning the agency in a community, sharing a recommendation with a colleague, or posting about it to an audience she never even knew existed.

You won't find that in a last-click attribution report. Likewise, you're probably never going to measure every recommendation.

Amanda's point is simple: the dashboard was never the whole picture.

Whether it's zero-click search, AI citations, or attribution, the common thread is the same. The old ways of measuring visibility are no longer enough on their own. They still matter, but they need to sit alongside a broader understanding of how people discover, evaluate, and recommend businesses today.

Thanks for reading. We hope you found this helpful, let us know on LinkedIn.

Content 10x is the world’s leading authority on content repurposing. Specializing in transforming your content into multiple formats and distributing it across various platforms, Content 10x helps maximize your reach and impact. Our expert team takes your core content and creates a diverse array of high-quality, engaging assets, including videos, podcasts, social media posts, and more. Trust Content 10x to enhance your content strategy and ensure your message is seen and heard by the right audience.

Be everywhere. Connect with more people. Grow your business.